Upload folder using huggingface_hub
Browse files- github-account-sdsc-ordes/github_account_github_project_metadata.json +11 -11
- github-account-sdsc-ordes/github_account_github_project_metadata.parquet +2 -2
- github-account-sdsc-ordes/github_account_github_repo_metadata.json +11 -11
- github-account-sdsc-ordes/github_account_github_repo_metadata.parquet +2 -2
github-account-sdsc-ordes/github_account_github_project_metadata.json
CHANGED
|
@@ -4321,9 +4321,9 @@
|
|
| 4321 |
"project_foreign_key": null,
|
| 4322 |
"pull_requests_closed": 0,
|
| 4323 |
"pull_requests_merged": 7,
|
| 4324 |
-
"pull_requests_open":
|
| 4325 |
-
"pull_requests_total":
|
| 4326 |
-
"pushed_at": "2026-
|
| 4327 |
"readme_length": 8525,
|
| 4328 |
"readme_text": "<p align=\"center\">\n <img src=\"./docs/assets/mava_logo.svg\" alt=\"project logo\" width=\"250\">\n</p>\n\n<h1 align=\"center\">\n mava-exchange\n</h1>\n<p align=\"center\">\n</p>\n\n\n[](https://github.com/sdsc-ordes/mava-exchange/actions/workflows/normal.yaml)\n[](http://www.apache.org/licenses/LICENSE-2.0.html)\n\n**Authors:**\n\n- [Stefan Milosavljevic](mailto:stefan.milosavljevic@sdsc.ethz.ch)\n- [Sabine Maennel](mailto:sabine.maennel@sdsc.ethz.ch)\n\n# mava-exchange\n\nA Python library and standard package format for exchanging video annotation\ndata between research tools.\n\n## Project Context\n\nThis work is part of the MAVA Project, which aims to improve interoperability\nand data exchange among three research tools — VideoScope, TIB-AV-A, and VIAN.\nBy standardising data formats and developing tools to import, export, and\nvalidate annotation packages, the project enhances data sharing and analysis\ncapabilities across linguistic, media studies, and audiovisual research.\n\nThis infrastructure will enable shared research workflows and ensure adherence\nto FAIR principles, improving the accessibility and reusability of research\ndata.\n\n## The Challenge\n\nVideo analysis via AI is computationally expensive and produces large datasets —\ncontinuous observations such as emotion scores, scene analysis, and audio\nfeatures accumulate quickly across a corpus. It is therefore desirable to share\nthese results among video processing tools without recomputing them.\n\nWhat is needed is a common data exchange format and tools to write, read, and\nvalidate packages in that format. Packages should be:\n\n- Interoperable and self-describing\n- Efficient to write and compact in file size\n\n## mava-exchange\n\n`mava-exchange` addresses this with a standard package format for video\nannotation corpora called `.mediapkg`. The library has two goals:\n\n**Format definition** — `mava-exchange` defines the `.mediapkg` standard: a ZIP\narchive containing annotation data as Parquet files alongside a manifest that\nmaps columns to the MAVA ontology via JSON-LD.\n\n**Tooling** — `mava-exchange` provides a Python library and CLI tools to write,\nread, inspect, and validate `.mediapkg` packages.\n\n## Design Choices\n\n- **ZIP + Parquet** — the `.mediapkg` is a ZIP archive containing one Parquet\n file per annotation track. Parquet offers compact storage, efficient reads,\n and columnar access. ZIP provides a universal container that any tool can open\n for inspection.\n\n- **JSON-LD manifest** — each package contains a `manifest.json` with a JSON-LD\n `@context` that maps Parquet column names to terms in the MAVA ontology. This\n is the semantic layer — it describes what each column means without being part\n of the data itself.\n\n- **MAVA ontology** — the ontology defines a shared vocabulary for annotation\n tracks, time coordinates, and observation dimensions. SHACL shapes are\n included for formal validation.\n\n- **Python** — the library is implemented in Python, as all participating tools\n use Python. Support for other languages may be added in future releases.\n\n## Inspiration\n\nThis format is inspired by GeoJSON and GeoParquet. GeoParquet embeds spatial\nmetadata inside Parquet files to describe geometry columns. `mava-exchange`\napplies the same principle to temporal data: where GeoParquet uses spatial\ncoordinates, `mava-exchange` uses time coordinates on a video timeline. Where\nGeoParquet metadata is purely operational, `mava-exchange` adds a semantic layer\nvia JSON-LD to link columns to a shared ontology.\n\n## Using the library\n\nInstall from PyPI:\n\n```bash\npip install mava-exchange\n# or with uv:\nuv add mava-exchange\n```\n\nWrite, read, and validate `.mediapkg` packages:\n\n```python\nfrom mava_exchange import (\n MediaPackageWriter, MediaPackageReader,\n ObservationSeries, AnnotationSeries, DimensionSpec,\n)\n\n# Define what your tracks mean\nemotions = ObservationSeries(\n name=\"emotions\",\n description=\"Face emotion scores from DeepFace\",\n sampling_interval=0.5,\n dimensions=[\n DimensionSpec(\"angry\", \"Anger probability\", \"[0,1]\"),\n DimensionSpec(\"neutral\", \"Neutral expression\", \"[0,1]\"),\n ]\n)\ntranscript = AnnotationSeries(\n name=\"transcript\",\n description=\"Whisper speech-to-text segments\",\n)\n\n# Write a package\nwith MediaPackageWriter(\"corpus.mediapkg\") as writer:\n writer.add_video(\"video_001\", \"https://example.org/talk.mp4\")\n writer.add_track(\"video_001\", emotions, emotions_df)\n writer.add_track(\"video_001\", transcript, transcript_df)\n\n# Read it back\nwith MediaPackageReader(\"corpus.mediapkg\") as reader:\n df = reader.read_track(\"video_001\", \"emotions\")\n```\n\nTwo CLI tools are also available after installation:\n\n```bash\nmediapkg-inspect corpus.mediapkg\nmediapkg-validate corpus.mediapkg\n```\n\nSee [`docs/tutorial.md`](docs/tutorial.md) for a complete walkthrough.\n\n## Development\n\nClone the repository and install in editable mode with development dependencies:\n\n```bash\ngit clone https://github.com/sdsc-ordes/mava-exchange.git\ncd mava-exchange\nuv sync --group dev\n```\n\nThe project uses [just](https://github.com/casey/just) as a task runner:\n\n```bash\njust test # run the test suite\njust lint # run ruff\njust format # format with ruff and treefmt\njust build # build the package\n```\n\nTo run the example that converts real TSV annotation files into a `.mediapkg`\ncorpus:\n\n```bash\njust example\njust inspect # inspect the resulting corpus.mediapkg\njust validate # validate it\n```\n\n## Further Reading\n\n- [`spec/SPEC.md`](spec/SPEC.md) — full format specification\n- [`spec/mava.ttl`](spec/mava.ttl) — MAVA ontology\n- [`spec/mava.shacl.ttl`](spec/mava.shacl.ttl) — MAVA shacl shapes\n- [`docs/tutorial.md`](docs/tutorial.md) — step-by-step guide\n- `examples`— complete example converting TSV files to `.mediapkg`\n\nFor development:\n\n- [Contribution Guidelines](/CONTRIBUTING.md)\n- [Development Guide](docs/development-guide.md)\n\n## Acknowledgements\n\nThis work was funded by the Swiss Data Science Center (SDSC) through its\nNational Call for Projects as an Infrastructure project.\n\nWe gratefully acknowledge the contributions of the SDSC experts and our\npartners.\n\n**SDSC Experts**:\n\n- Dr. Stefan Milosavljevic, ORCID ID\n [0000-0002-9135-1353](https://orcid.org/0000-0002-9135-1353)\n- Sabine Maennel, ORCID ID\n [0009-0001-3022-8239](https://orcid.org/0009-0001-3022-8239)\n- Robin Franken, ORCID ID\n [0009-0008-0143-9118](https://orcid.org/0009-0008-0143-9118)\n- Dr. Oksana Riba Grognuz, ORCID ID\n [0000-0002-2961-2655](https://orcid.org/0000-0002-2961-2655)\n\n**Partner Institutions**\n\n- Dr. Teodora Vuković, ORCID ID\n [0000-0002-5780-5665](https://orcid.org/0000-0002-5780-5665)\n- Dr. Jeremy Zehr, ORCID ID\n [0000-0002-6046-8647](https://orcid.org/0000-0002-6046-8647)\n- Prof. Dr. Josephine Diecke, ORCID ID\n [0000-0002-9342-0631](https://orcid.org/0000-0002-9342-0631)\n- Dr. Simon Spiegel, ORCID ID\n [0000-0003-2141-5566](https://orcid.org/0000-0003-2141-5566)\n- Prof. Dr. Ralph Ewerth, ORCID ID\n [0000-0003-0918-6297](https://orcid.org/0000-0003-0918-6297)\n- Dr. Eric Müller-Budack, ORCID ID\n [0000-0002-6802-1241](https://orcid.org/0000-0002-6802-1241)\n- Dr. Cristina Grisot, ORCID ID\n [0000-0003-0684-4442](https://orcid.org/0000-0003-0684-4442)\n\n## How to Cite\n\nIf you use this software, please cite it as follows:\n\n👉 See the [CITATION.cff](./CITATION.cff) file for the full list of software\nauthors and citation formats.\n\nWhen referring to the project more broadly (including partner contributions),\nplease acknowledge the funding statement and collaborators listed in the\n[Acknowledgements](#acknowledgements) section:\n\n> \"This work was funded by the Swiss Data Science Center (SDSC) through its\n> National Call for Projects as an Infrastructure project.\"\n\n## Copyright\n\nCopyright © 2025-2026 Swiss Data Science Center\n(SDSC),[www.datascience.ch](http://www.datascience.ch/), ROR:\n[ror.org/02hdt9m26](https://ror.org/02hdt9m26). All rights reserved. The SDSC is\na Swiss National Research Infrastructure, jointly established and legally\nrepresented by the École Polytechnique Fédérale de Lausanne (EPFL) and the\nEidgenössische Technische Hochschule Zürich (ETH Zürich) as a société simple.\nThis copyright encompasses all materials, software, documentation, and other\ncontent created and developed by the SDSC.",
|
| 4329 |
"readme_title": "mava-exchange",
|
|
@@ -7339,7 +7339,7 @@
|
|
| 7339 |
"pull_requests_merged": 2,
|
| 7340 |
"pull_requests_open": 1,
|
| 7341 |
"pull_requests_total": 3,
|
| 7342 |
-
"pushed_at": "2026-03-
|
| 7343 |
"readme_length": 2106,
|
| 7344 |
"readme_text": "<p align=\"center\">\n <img src=\"./docs/assets/logo.svg\" alt=\"project logo\" width=\"250\">\n</p>\n\n<h1 align=\"center\">\n open-pulse-hackathon-analysis\n</h1>\n<p align=\"center\">\n</p>\n\n[](https://github.com/sdsc-ordes/open-pulse-hackathon-analysis/releases/latest)\n[](https://github.com/sdsc-ordes/open-pulse-hackathon-analysis/actions/workflows/normal.yaml)\n\n**Authors:**\n\n- [Eisha Tir Raazia](mailto:eisha.raazia@epfl.ch)\n\n## Installation\n\nDescribe the installation instruction here.\n\n## Usage\n\n### Extract Hackathon Data\n\nExtract data from LauzHack:\n\n```bash\njust hackalysis --output_folder ./data --hackathon_provider lauzhack\n```\n\nExtract specific LauzHack years with the CLI:\n\n```bash\nuv run hackalysis -o ./data -p lauzhack -y 2023,2024,2025\n```\n\n````\nExtract data without specifying the years will default to extracting all available years for the specified provider.\n```bash\nuv run hackalysis -o ./data -p lauzhack\n```\n\nExtract data from Devpost:\n\n```bash\njust hackalysis --output_folder ./data --hackathon_provider devpost --hackathon_name \"ExampleHackathonName\"\n```\nor\n\n```bash\njust hackalysis -o ./data -p devpost -n \"ExampleHackathonName\"\n```\n\n## Development\n\nRead first the [Contribution Guidelines](/CONTRIBUTING.md).\n\nFor technical documentation on setup and development, see the\n[Development Guide](docs/development-guide.md)\n\n## Acknowledgement\n\nAcknowledge all contributors and external collaborators here.\n\n## Copyright\n\nCopyright © 2026-2028 Swiss Data Science Center (SDSC),\n[www.datascience.ch](http://www.datascience.ch/). All rights reserved. The SDSC\nis jointly established and legally represented by the École Polytechnique\nFédérale de Lausanne (EPFL) and the Eidgenössische Technische Hochschule Zürich\n(ETH Zürich). This copyright encompasses all materials, software, documentation,\nand other content created and developed by the SDSC.",
|
| 7345 |
"readme_title": "<p align=\"center\">",
|
|
@@ -8324,11 +8324,11 @@
|
|
| 8324 |
],
|
| 8325 |
"github_repos_metadata": [
|
| 8326 |
{
|
| 8327 |
-
"commit_count_default_branch":
|
| 8328 |
"contributors_count": 2,
|
| 8329 |
"contributors_top": [
|
| 8330 |
{
|
| 8331 |
-
"contributions":
|
| 8332 |
"html_url": "https://github.com/gabyx",
|
| 8333 |
"login": "gabyx"
|
| 8334 |
},
|
|
@@ -8471,7 +8471,7 @@
|
|
| 8471 |
"languages_top": [
|
| 8472 |
{
|
| 8473 |
"language": "Go",
|
| 8474 |
-
"size":
|
| 8475 |
},
|
| 8476 |
{
|
| 8477 |
"language": "Nix",
|
|
@@ -8490,8 +8490,8 @@
|
|
| 8490 |
"size": 524
|
| 8491 |
}
|
| 8492 |
],
|
| 8493 |
-
"last_commit_date_default_branch": "2026-03-
|
| 8494 |
-
"last_commit_oid_default_branch": "
|
| 8495 |
"latest_release_date": "2026-03-12T16:51:31Z",
|
| 8496 |
"latest_release_tag": "v0.39.0",
|
| 8497 |
"license_name": "MIT License",
|
|
@@ -8506,7 +8506,7 @@
|
|
| 8506 |
"pull_requests_merged": 59,
|
| 8507 |
"pull_requests_open": 0,
|
| 8508 |
"pull_requests_total": 61,
|
| 8509 |
-
"pushed_at": "2026-03-
|
| 8510 |
"readme_length": 14597,
|
| 8511 |
"readme_text": "<p align=\"center\">\n <img src=\"docs/assets/images/logo.svg\" width=\"400px\">\n</p>\n\n<p align=\"center\">\n <a href=\"https://codecov.io/gh/sdsc-ordes/quitsh\">\n <img src=\"https://codecov.io/gh/sdsc-ordes/quitsh/graph/badge.svg?token=35RP18MHFR\" alt=\"Coverage\" />\n </a>\n <a href=\"https://github.com/sdsc-ordes/quitsh/releases/latest\">\n <img src=\"https://img.shields.io/github/release/sdsc-ordes/quitsh.svg?label=release\" alt=\"Current Release\" />\n </a>\n <a href=\"https://pkg.go.dev/github.com/sdsc-ordes/quitsh\">\n <img src=\"https://pkg.go.dev/badge/github.com/sdsc-ordes/quitsh\" alt=\"Package\" />\n </a>\n <a href=\"https://github.com/sdsc-ordes/quitsh/actions/workflows/normal.yaml\">\n <img src=\"https://img.shields.io/github/actions/workflow/status/sdsc-ordes/quitsh/normal.yaml?label=ci\" alt=\"Pipeline Status\" />\n </a>\n <a href=\"https://mit-license.org/\">\n <img src=\"https://img.shields.io/badge/License-Apache2.0-blue.svg?\" alt=\"License label\" />\n </a>\n</p>\n\n---\n\n> [!CAUTION]\n>\n> This repository is in `beta`. The design space of this tool is still explored\n> and breaking changes might occure. Documentation is incomplete.\n\nThe `quitsh` framework (`/kwɪʧ/`) is a build-tooling **CLI library/framework**\ndesigned to replace loosely-typed scripting languages (e.g., `bash`, `python`,\nand similar alternatives) with the statically-typed language `Go`. Its goal is\nto simplify tooling tasks while providing robust, extendable solutions for\ncomponent repositories (mono-repositories).\n\n`quitsh` is an opinionated framework born out of frustration with the lack of\nsimple and extendable tooling for mono-repos. It is language-agnostic and\ntoolchain-independent, allowing users to focus on their workflows without being\nconstrained by specific technologies.\n\nQuitsh can be used in two non-coupled ways (enhancing each other):\n\n- Build/use & extend the CLI `quitsh` for your tooling/CI/CD scripts in your\n repository. You write `quitsh mycommand --do-it` and decide what it does,\n using the [library modules](#built-in-libraries) for running subprocesses,\n logging etc.\n\n- Use its [component feature](#component-system) which gives you the ability to\n register your tooling scripts (build/lint/test etc.) as\n [runners](#runner-system) which can be reused across\n [components](#components). Since runners depend heavily on available\n executables in your `PATH`, runners run over a toolchain (currently Nix\n development shell etc.).\n\n## Key Features\n\n### Code-First Approach\n\n- All tooling logic is implemented in `Go`.\n- Tooling logic is defined primarily in code, avoiding declarative\n configurations or templated non-typed languages, which often add unnecessary\n complexity despite their flexibility. _Note: Supporting configuration\n languages like `jsonnet` or \"turing-complete\" YAML etc. is particularly\n avoided. **Everything stays strongly-typed and fast compiled.**_\n\n### Extendability\n\n- `quitsh` serves as a library to build **your customized** CLI tool for your\n specific tasks.\n- Users can add custom commands and specialized tooling features using libraries\n like [`cobra`](https://github.com/spf13/cobra), `quitsh mycommand --do-it`.\n\n### Component System\n\n- [Components](#components) (i.e., buildable units) are identified by placing a\n configuration file (default: `.component.yaml`) in the corresponding\n subdirectory.\n\n#### Targets and Steps\n\n- Each component defines **targets**, which consist of multiple **steps**.\n- Targets can depend on other targets across the repository.\n- Input change sets can be specified for each target to track modifications and\n determine if the target is outdated.\n\n#### Runner System\n\n- Steps within targets are executed by **runners**, which are written by you in\n Go and act as reusable replacements for traditional build/tooling scripts.\n- Runners can have custom YAML configuration options specified per component in\n `.component.yaml`.\n\n#### Toolchain Dispatch\n\n- Runners are associated with specific toolchains.\n- By default, `quitsh` includes a\n [Nix development shell dispatch](https://nix.dev/tutorials/first-steps/declarative-shell.html),\n providing stable and reproducible environments.\n- While container-based dispatching is not a primary goal, it can be implemented\n by extending the dispatch interface.\n\n- The tool was built to replicate the same procedure one executes during local\n development and also in CI. Having CI align with what you execute locally is\n not a nice thing to have, its a necessity. Nix development shells (or\n containers) help with this. A Nix shell provides a simple and robust\n abstraction to pin a toolchain. The following visualization gives an overview\n about how `quitsh` is used:\n\n <p align=\"center\">\n\n \n\n </p>\n\n### Built-in Libraries\n\nThe `pkg` folder offers utilities for common development needs, such as:\n\n- Command Execution: [`pkg/exec`](pkg/exec) provides utilities for process\n execution and command chaining.\n- Structured Logging: [`pkg/log`](pkg/log) enables consistent and readable\n logging.\n- Error Handling: [`pkg/error`](pkg/error) facilitates contextual error\n management.\n- Dependency Graphs: Tools for managing and resolving dependency graphs across\n targets.\n- Some Go `test` runners (here as an example) for running Go tests (its used\n internally to test `quitsh` it-self).\n\n#### Performance\n\n- Since all tooling is written in `Go`, `quitsh` provides type safety and fast\n performance by default.\n- Combined with a Nix-based toolchain dispatch and the ability to write tests\n easily, the framework significantly accelerates the \"change, test, improve\"\n workflow.\n\n#### Nix Integration\n\n- A CLI tool built with `quitsh` can be\n [seamlessly packaged](./tools/nix/packages/cli) into a Nix development shells,\n ensuring accessibility for all users of a given repository.\n\n---\n\n# How To Use It?\n\nUsing this library follows instantiating the CLI (also demonstrated in this\nrepository in [`main.go`](./tools/cli/cmd/cli/main.go), e.g.:\n\n```go\nargs := cliconfig.New()\n\ncli, err := cli.New(\n &args.Commands.Root,\n &args,\n cli.WithName(\"cli\"),\n cli.WithDescription(\"This is the 🐔-🥚 CLI tool for 'quitsh', yes its built with 'quitsh'.\"),\n cli.WithCompFindOptions(\n query.WithFindOptions(\n fs.WithWalkDirFilterPatterns(nil,\n []string{\"**/test/repo/**\"}, true))),\n cli.WithStages(\"lint\", \"build\", \"test\"),\n cli.WithTargetToStageMapperDefault(),\n cli.WithToolchainDispatcherNix(\n \"tools/nix\",\n func(c config.IConfig) *toolchain.DispatchArgs {\n cc := common.Cast[*cliconfig.Config](c)\n\n return &cc.Commands.DispatchArgs\n },\n ),\n)\n```\n\nYou can now add runners and your own commands depending on the needs of your\nrepository. For example in [`main.go`](./tools/cli/cmd/cli/main.go):\n\n```go\nlistcmd.AddCmd(cli, cli.RootCmd())\nconfigcmd.AddCmd(cli.RootCmd(), &conf)\nexectarget.AddCmd(cli, cli.RootCmd())\nexecrunner.AddCmd(cli, cli.RootCmd(), &conf.Commands.DispatchArgs)\n```\n\nadds essential `quitsh` commands\n\n- `listcmd` to list all components etc: `quitsh list`.\n- `configcmd` to inspect/write the config file: `quitsh config ...`.\n- `exectarget` to execute specific targets `quitsh exec-target`.\n- `execrunner` to let `quitsh` dispatch over toolchains (see\n `cli.WithToolchainDispatcherNix` above): `quitsh exec-runner ...`\n\nThere are lots of more useful commands in [`pkg/cli/cmd`](./pkg/cli/cmd) which\nyou might use.\n\n## Useful References\n\nA reference repository with commands and runner can be looked at here:\n\n- [`Custodian`](https://gitlab.com/data-custodian/custodian/tree/main/tools/quitsh)\n- [`DAC-Portal`](https://gitlab.com/data-custodian/dac-portal/tree/main/tools/quitsh)\n- [`Quitsh (this repo)`](https://github.com/sdsc-ordes/quitsh)\n\n## Config\n\nQuitsh runs with global config YAML file which it loads (or defaults) at startup\nfor any invocation. The above CLI instantiation constructs a new config with\n`cliconfig.New()` (this is custom for each usecase and can be adjusted and\nmodified). The config defines global settings (output directories, logging etc.)\nand also various custom, use-case specific settings. These might include\nsettings which runners (or custom commands) might use during execution. For\nexample build runners might use a\n[`build.BuildType`](./tools/cli/pkg/config/config.go) property which could be\n`debug` or `release` etc. The CLI does not care about your custom settings, they\nonly need to be fully serializable to YAML (for toolchain dispatching) and you\ncan override defaults from custom added commands for example.\n\n### Modifying Config Values\n\nYou have the ability to set the config file `quitsh` uses with `--config` or\nread it from stdin with `--config -` or set options (YAML) on the command line\nwith `--config-value`. For example `--config-value \"build.buildType: release\"`\nwould set the `build.BuildType` setting to `release` on startup.\n\n## Components\n\nQuitsh builds around **components**. A **component** should be treated as an own\ncompartment/Git repository or directory with its own independent source code and\noutput.\n\nA component in `quitsh` is defined by a `.component.yaml` (name is\nconfigurable):\n\n```yaml\n# The name of the component: Must be a unique.\nname: my-component\n\n# A semantic version.\n# This is useful for Nix packaging etc.\nversion: 0.2.1\n\n# A simple annotation (not used internally) what main language this component uses.\nlanguage: go\n\n# The `.general` object is not parsed by `quitsh` and\n# allows arbitrary values mainly used for YAML anchors.\n.general:\n value: &val 60\n\ntargets:\n # A target called `test` with two steps.\n my-test:\n # The stage to which this target belongs. Does not need to be provided\n # if the CLI is setup to map target names to stages.\n stage: test\n\n steps:\n # Step 1: Using runner with ID (how it was registered).\n - runner-id: banana-project::my-test-runner\n config: # Your custom runner YAML config, (optional).\n\n # Step 2: Using a runner with registered key (stage: `test`, name `my-test`)\n - runner: my-test\n\n # A target called `build-all` with one step.\n build-all:\n stage: build\n\n # Defining when this target is considered changed:\n # i.e. whenever `self::sources` input change set is changed.\n # `self` maps to this component.\n inputs: [\"self::sources\"]\n\n # Defining dependencies on other targets such that this\n # target is executed after target `my-test` above.\n # You can also link to other components (e.g `other-comp::build`).\n depends: [\"self::my-test\"]\n\n steps:\n # Step 1: Using a runner with registered key (stage: `build`, name `my-test`)\n - runner: my-build\n config:\n tags: [\"doit\"]\n\n lint:\n steps:\n - ... other steps ...\n\ninputs:\n # An input change set with name `sources` which defines\n # patterns to match all source files.\n sources:\n # A regex which matches `*.go` files in `./src` in the components folder.\n patterns:\n - '^./src/.*\\.go$'\n```\n\n## Execution of Targets\n\nThe execution of steps by `quitsh` is done by reading a\n[`.component.yaml`](.component.yaml) for each component. The\n[`.component.yaml`](.component.yaml) file contains _inputs_ and _targets_.\n\nQuitsh's own [`.component.yaml`](./.component.yaml) looks like:\n\n```yaml\nname: quitsh\nlanguage: go\n\ninputs:\n srcs:\n patterns:\n - \"^.*$\"\n\ntargets:\n test:\n stage: test\n steps:\n - runner: go\n - runner: go-bin\n config:\n # Build everything instrumented.\n # Execute the binaries via a `go test` provided in the following\n # pkg and with tags.\n buildPkg: test/cmd\n testPkg: test\n buildTags: [\"integration\"]\n testTags: [\"integration\"]\n\n build:\n steps:\n - runner: go\n lint:\n steps:\n - runner: go\n```\n\nEach **target** defines is a set of **steps** which itself are further specified\nby the field `runner`. A **runner** is **Go code** applicable for a certain step\nwhich should work for all components.\n\nA runner is registered in [`factory`](./pkg/runner/factory/runner.go), for\nexample [here](./tools/cli/pkg/runner/go/lint.go). Runners can be written by\nimplementing the interface [`Runner`](./pkg/runner/runner.go) inside\n[`./pkg/runner/runners`](./pkg/runner/runners) and registering them in\n[`./pkg/runner/factory/init-runners.go`], for example\n[here](./tools/cli/pkg/runner/go/register.go).\n\n> [!NOTE]\n>\n> **You can execute targets in parallel with `--parallel`**.\n\n## Runner Configuration\n\nRunners can load independent YAML config under `config` to make them\nconfigurable, e.g. the `go` build runner loads the following\n[config](./pkg/runner/runners/go/build-config.go):\n\n```yaml\nsteps:\n build:\n - runner: go\n config:\n version-module: \"pkg/myversion-module\" # defaults to `pkg/build`\n```\n\n## Target Stages\n\nEach target also maps to a _stage_ which `quitsh` uses to group targets together\nif you want to find them and make gathering commands such as the example\n[`build` here](https://gitlab.com/data-custodian/custodian/-/blob/main/tools/quitsh/cmd/quitsh/cmd/build/build.go#L83).\nIt collects and runs all targets in the stage `build`.\n\n## How to Extend Functionality\n\nFor users using the `quitsh` CLI framework, it is suggested to follow the\nfollowing points when thinking about new functionality in a repository which\nuses this library. If you need new functionality for CI and local development\nwhich you normally would write in `bash`/`python` follow the following steps:\n\n- If the functionality is **a feature needed in an existing runner and step**:\n Extend the runner and make it work with your new test/build/lint feature.\n\n- If the functionality is **not related to a runner or the same for each\n component with that language**: Extend `quitsh` by providing another\n subcommand which does what you need, see this example\n [`fix-hash`](https://gitlab.com/data-custodian/custodian/-/blob/main/tools/quitsh/cmd/quitsh/cmd/nix/fix-hash/fix.go).\n\n- If the functionality is **for a certain language, e.g. `go` or `python` and\n applies to each component which is written in that language**: consider adding\n a new runner for an already pre-defined stage, e.g. `lint`, `build` etc.\n\n## Example Applications\n\nUnderstand what this framework does, is best accomplished by understanding how\nwe use this framework in our\n[components repo (mono-repo)](https://gitlab.com/data-custodian/custodian). Our\nmajor components are located in\n[./components](https://gitlab.com/data-custodian/custodian/-/tree/main/components).",
|
| 8512 |
"readme_title": "<p align=\"center\">",
|
|
@@ -8525,7 +8525,7 @@
|
|
| 8525 |
"monorepo",
|
| 8526 |
"cli"
|
| 8527 |
],
|
| 8528 |
-
"updated_at": "2026-03-
|
| 8529 |
"url": "https://github.com/sdsc-ordes/quitsh",
|
| 8530 |
"watchers": 1
|
| 8531 |
}
|
|
|
|
| 4321 |
"project_foreign_key": null,
|
| 4322 |
"pull_requests_closed": 0,
|
| 4323 |
"pull_requests_merged": 7,
|
| 4324 |
+
"pull_requests_open": 2,
|
| 4325 |
+
"pull_requests_total": 9,
|
| 4326 |
+
"pushed_at": "2026-03-16T13:25:15Z",
|
| 4327 |
"readme_length": 8525,
|
| 4328 |
"readme_text": "<p align=\"center\">\n <img src=\"./docs/assets/mava_logo.svg\" alt=\"project logo\" width=\"250\">\n</p>\n\n<h1 align=\"center\">\n mava-exchange\n</h1>\n<p align=\"center\">\n</p>\n\n\n[](https://github.com/sdsc-ordes/mava-exchange/actions/workflows/normal.yaml)\n[](http://www.apache.org/licenses/LICENSE-2.0.html)\n\n**Authors:**\n\n- [Stefan Milosavljevic](mailto:stefan.milosavljevic@sdsc.ethz.ch)\n- [Sabine Maennel](mailto:sabine.maennel@sdsc.ethz.ch)\n\n# mava-exchange\n\nA Python library and standard package format for exchanging video annotation\ndata between research tools.\n\n## Project Context\n\nThis work is part of the MAVA Project, which aims to improve interoperability\nand data exchange among three research tools — VideoScope, TIB-AV-A, and VIAN.\nBy standardising data formats and developing tools to import, export, and\nvalidate annotation packages, the project enhances data sharing and analysis\ncapabilities across linguistic, media studies, and audiovisual research.\n\nThis infrastructure will enable shared research workflows and ensure adherence\nto FAIR principles, improving the accessibility and reusability of research\ndata.\n\n## The Challenge\n\nVideo analysis via AI is computationally expensive and produces large datasets —\ncontinuous observations such as emotion scores, scene analysis, and audio\nfeatures accumulate quickly across a corpus. It is therefore desirable to share\nthese results among video processing tools without recomputing them.\n\nWhat is needed is a common data exchange format and tools to write, read, and\nvalidate packages in that format. Packages should be:\n\n- Interoperable and self-describing\n- Efficient to write and compact in file size\n\n## mava-exchange\n\n`mava-exchange` addresses this with a standard package format for video\nannotation corpora called `.mediapkg`. The library has two goals:\n\n**Format definition** — `mava-exchange` defines the `.mediapkg` standard: a ZIP\narchive containing annotation data as Parquet files alongside a manifest that\nmaps columns to the MAVA ontology via JSON-LD.\n\n**Tooling** — `mava-exchange` provides a Python library and CLI tools to write,\nread, inspect, and validate `.mediapkg` packages.\n\n## Design Choices\n\n- **ZIP + Parquet** — the `.mediapkg` is a ZIP archive containing one Parquet\n file per annotation track. Parquet offers compact storage, efficient reads,\n and columnar access. ZIP provides a universal container that any tool can open\n for inspection.\n\n- **JSON-LD manifest** — each package contains a `manifest.json` with a JSON-LD\n `@context` that maps Parquet column names to terms in the MAVA ontology. This\n is the semantic layer — it describes what each column means without being part\n of the data itself.\n\n- **MAVA ontology** — the ontology defines a shared vocabulary for annotation\n tracks, time coordinates, and observation dimensions. SHACL shapes are\n included for formal validation.\n\n- **Python** — the library is implemented in Python, as all participating tools\n use Python. Support for other languages may be added in future releases.\n\n## Inspiration\n\nThis format is inspired by GeoJSON and GeoParquet. GeoParquet embeds spatial\nmetadata inside Parquet files to describe geometry columns. `mava-exchange`\napplies the same principle to temporal data: where GeoParquet uses spatial\ncoordinates, `mava-exchange` uses time coordinates on a video timeline. Where\nGeoParquet metadata is purely operational, `mava-exchange` adds a semantic layer\nvia JSON-LD to link columns to a shared ontology.\n\n## Using the library\n\nInstall from PyPI:\n\n```bash\npip install mava-exchange\n# or with uv:\nuv add mava-exchange\n```\n\nWrite, read, and validate `.mediapkg` packages:\n\n```python\nfrom mava_exchange import (\n MediaPackageWriter, MediaPackageReader,\n ObservationSeries, AnnotationSeries, DimensionSpec,\n)\n\n# Define what your tracks mean\nemotions = ObservationSeries(\n name=\"emotions\",\n description=\"Face emotion scores from DeepFace\",\n sampling_interval=0.5,\n dimensions=[\n DimensionSpec(\"angry\", \"Anger probability\", \"[0,1]\"),\n DimensionSpec(\"neutral\", \"Neutral expression\", \"[0,1]\"),\n ]\n)\ntranscript = AnnotationSeries(\n name=\"transcript\",\n description=\"Whisper speech-to-text segments\",\n)\n\n# Write a package\nwith MediaPackageWriter(\"corpus.mediapkg\") as writer:\n writer.add_video(\"video_001\", \"https://example.org/talk.mp4\")\n writer.add_track(\"video_001\", emotions, emotions_df)\n writer.add_track(\"video_001\", transcript, transcript_df)\n\n# Read it back\nwith MediaPackageReader(\"corpus.mediapkg\") as reader:\n df = reader.read_track(\"video_001\", \"emotions\")\n```\n\nTwo CLI tools are also available after installation:\n\n```bash\nmediapkg-inspect corpus.mediapkg\nmediapkg-validate corpus.mediapkg\n```\n\nSee [`docs/tutorial.md`](docs/tutorial.md) for a complete walkthrough.\n\n## Development\n\nClone the repository and install in editable mode with development dependencies:\n\n```bash\ngit clone https://github.com/sdsc-ordes/mava-exchange.git\ncd mava-exchange\nuv sync --group dev\n```\n\nThe project uses [just](https://github.com/casey/just) as a task runner:\n\n```bash\njust test # run the test suite\njust lint # run ruff\njust format # format with ruff and treefmt\njust build # build the package\n```\n\nTo run the example that converts real TSV annotation files into a `.mediapkg`\ncorpus:\n\n```bash\njust example\njust inspect # inspect the resulting corpus.mediapkg\njust validate # validate it\n```\n\n## Further Reading\n\n- [`spec/SPEC.md`](spec/SPEC.md) — full format specification\n- [`spec/mava.ttl`](spec/mava.ttl) — MAVA ontology\n- [`spec/mava.shacl.ttl`](spec/mava.shacl.ttl) — MAVA shacl shapes\n- [`docs/tutorial.md`](docs/tutorial.md) — step-by-step guide\n- `examples`— complete example converting TSV files to `.mediapkg`\n\nFor development:\n\n- [Contribution Guidelines](/CONTRIBUTING.md)\n- [Development Guide](docs/development-guide.md)\n\n## Acknowledgements\n\nThis work was funded by the Swiss Data Science Center (SDSC) through its\nNational Call for Projects as an Infrastructure project.\n\nWe gratefully acknowledge the contributions of the SDSC experts and our\npartners.\n\n**SDSC Experts**:\n\n- Dr. Stefan Milosavljevic, ORCID ID\n [0000-0002-9135-1353](https://orcid.org/0000-0002-9135-1353)\n- Sabine Maennel, ORCID ID\n [0009-0001-3022-8239](https://orcid.org/0009-0001-3022-8239)\n- Robin Franken, ORCID ID\n [0009-0008-0143-9118](https://orcid.org/0009-0008-0143-9118)\n- Dr. Oksana Riba Grognuz, ORCID ID\n [0000-0002-2961-2655](https://orcid.org/0000-0002-2961-2655)\n\n**Partner Institutions**\n\n- Dr. Teodora Vuković, ORCID ID\n [0000-0002-5780-5665](https://orcid.org/0000-0002-5780-5665)\n- Dr. Jeremy Zehr, ORCID ID\n [0000-0002-6046-8647](https://orcid.org/0000-0002-6046-8647)\n- Prof. Dr. Josephine Diecke, ORCID ID\n [0000-0002-9342-0631](https://orcid.org/0000-0002-9342-0631)\n- Dr. Simon Spiegel, ORCID ID\n [0000-0003-2141-5566](https://orcid.org/0000-0003-2141-5566)\n- Prof. Dr. Ralph Ewerth, ORCID ID\n [0000-0003-0918-6297](https://orcid.org/0000-0003-0918-6297)\n- Dr. Eric Müller-Budack, ORCID ID\n [0000-0002-6802-1241](https://orcid.org/0000-0002-6802-1241)\n- Dr. Cristina Grisot, ORCID ID\n [0000-0003-0684-4442](https://orcid.org/0000-0003-0684-4442)\n\n## How to Cite\n\nIf you use this software, please cite it as follows:\n\n👉 See the [CITATION.cff](./CITATION.cff) file for the full list of software\nauthors and citation formats.\n\nWhen referring to the project more broadly (including partner contributions),\nplease acknowledge the funding statement and collaborators listed in the\n[Acknowledgements](#acknowledgements) section:\n\n> \"This work was funded by the Swiss Data Science Center (SDSC) through its\n> National Call for Projects as an Infrastructure project.\"\n\n## Copyright\n\nCopyright © 2025-2026 Swiss Data Science Center\n(SDSC),[www.datascience.ch](http://www.datascience.ch/), ROR:\n[ror.org/02hdt9m26](https://ror.org/02hdt9m26). All rights reserved. The SDSC is\na Swiss National Research Infrastructure, jointly established and legally\nrepresented by the École Polytechnique Fédérale de Lausanne (EPFL) and the\nEidgenössische Technische Hochschule Zürich (ETH Zürich) as a société simple.\nThis copyright encompasses all materials, software, documentation, and other\ncontent created and developed by the SDSC.",
|
| 4329 |
"readme_title": "mava-exchange",
|
|
|
|
| 7339 |
"pull_requests_merged": 2,
|
| 7340 |
"pull_requests_open": 1,
|
| 7341 |
"pull_requests_total": 3,
|
| 7342 |
+
"pushed_at": "2026-03-16T09:09:15Z",
|
| 7343 |
"readme_length": 2106,
|
| 7344 |
"readme_text": "<p align=\"center\">\n <img src=\"./docs/assets/logo.svg\" alt=\"project logo\" width=\"250\">\n</p>\n\n<h1 align=\"center\">\n open-pulse-hackathon-analysis\n</h1>\n<p align=\"center\">\n</p>\n\n[](https://github.com/sdsc-ordes/open-pulse-hackathon-analysis/releases/latest)\n[](https://github.com/sdsc-ordes/open-pulse-hackathon-analysis/actions/workflows/normal.yaml)\n\n**Authors:**\n\n- [Eisha Tir Raazia](mailto:eisha.raazia@epfl.ch)\n\n## Installation\n\nDescribe the installation instruction here.\n\n## Usage\n\n### Extract Hackathon Data\n\nExtract data from LauzHack:\n\n```bash\njust hackalysis --output_folder ./data --hackathon_provider lauzhack\n```\n\nExtract specific LauzHack years with the CLI:\n\n```bash\nuv run hackalysis -o ./data -p lauzhack -y 2023,2024,2025\n```\n\n````\nExtract data without specifying the years will default to extracting all available years for the specified provider.\n```bash\nuv run hackalysis -o ./data -p lauzhack\n```\n\nExtract data from Devpost:\n\n```bash\njust hackalysis --output_folder ./data --hackathon_provider devpost --hackathon_name \"ExampleHackathonName\"\n```\nor\n\n```bash\njust hackalysis -o ./data -p devpost -n \"ExampleHackathonName\"\n```\n\n## Development\n\nRead first the [Contribution Guidelines](/CONTRIBUTING.md).\n\nFor technical documentation on setup and development, see the\n[Development Guide](docs/development-guide.md)\n\n## Acknowledgement\n\nAcknowledge all contributors and external collaborators here.\n\n## Copyright\n\nCopyright © 2026-2028 Swiss Data Science Center (SDSC),\n[www.datascience.ch](http://www.datascience.ch/). All rights reserved. The SDSC\nis jointly established and legally represented by the École Polytechnique\nFédérale de Lausanne (EPFL) and the Eidgenössische Technische Hochschule Zürich\n(ETH Zürich). This copyright encompasses all materials, software, documentation,\nand other content created and developed by the SDSC.",
|
| 7345 |
"readme_title": "<p align=\"center\">",
|
|
|
|
| 8324 |
],
|
| 8325 |
"github_repos_metadata": [
|
| 8326 |
{
|
| 8327 |
+
"commit_count_default_branch": 217,
|
| 8328 |
"contributors_count": 2,
|
| 8329 |
"contributors_top": [
|
| 8330 |
{
|
| 8331 |
+
"contributions": 216,
|
| 8332 |
"html_url": "https://github.com/gabyx",
|
| 8333 |
"login": "gabyx"
|
| 8334 |
},
|
|
|
|
| 8471 |
"languages_top": [
|
| 8472 |
{
|
| 8473 |
"language": "Go",
|
| 8474 |
+
"size": 414373
|
| 8475 |
},
|
| 8476 |
{
|
| 8477 |
"language": "Nix",
|
|
|
|
| 8490 |
"size": 524
|
| 8491 |
}
|
| 8492 |
],
|
| 8493 |
+
"last_commit_date_default_branch": "2026-03-16T08:16:51Z",
|
| 8494 |
+
"last_commit_oid_default_branch": "fcb9abac7c966f92ffe91dc496dc7506020d1280",
|
| 8495 |
"latest_release_date": "2026-03-12T16:51:31Z",
|
| 8496 |
"latest_release_tag": "v0.39.0",
|
| 8497 |
"license_name": "MIT License",
|
|
|
|
| 8506 |
"pull_requests_merged": 59,
|
| 8507 |
"pull_requests_open": 0,
|
| 8508 |
"pull_requests_total": 61,
|
| 8509 |
+
"pushed_at": "2026-03-16T08:17:48Z",
|
| 8510 |
"readme_length": 14597,
|
| 8511 |
"readme_text": "<p align=\"center\">\n <img src=\"docs/assets/images/logo.svg\" width=\"400px\">\n</p>\n\n<p align=\"center\">\n <a href=\"https://codecov.io/gh/sdsc-ordes/quitsh\">\n <img src=\"https://codecov.io/gh/sdsc-ordes/quitsh/graph/badge.svg?token=35RP18MHFR\" alt=\"Coverage\" />\n </a>\n <a href=\"https://github.com/sdsc-ordes/quitsh/releases/latest\">\n <img src=\"https://img.shields.io/github/release/sdsc-ordes/quitsh.svg?label=release\" alt=\"Current Release\" />\n </a>\n <a href=\"https://pkg.go.dev/github.com/sdsc-ordes/quitsh\">\n <img src=\"https://pkg.go.dev/badge/github.com/sdsc-ordes/quitsh\" alt=\"Package\" />\n </a>\n <a href=\"https://github.com/sdsc-ordes/quitsh/actions/workflows/normal.yaml\">\n <img src=\"https://img.shields.io/github/actions/workflow/status/sdsc-ordes/quitsh/normal.yaml?label=ci\" alt=\"Pipeline Status\" />\n </a>\n <a href=\"https://mit-license.org/\">\n <img src=\"https://img.shields.io/badge/License-Apache2.0-blue.svg?\" alt=\"License label\" />\n </a>\n</p>\n\n---\n\n> [!CAUTION]\n>\n> This repository is in `beta`. The design space of this tool is still explored\n> and breaking changes might occure. Documentation is incomplete.\n\nThe `quitsh` framework (`/kwɪʧ/`) is a build-tooling **CLI library/framework**\ndesigned to replace loosely-typed scripting languages (e.g., `bash`, `python`,\nand similar alternatives) with the statically-typed language `Go`. Its goal is\nto simplify tooling tasks while providing robust, extendable solutions for\ncomponent repositories (mono-repositories).\n\n`quitsh` is an opinionated framework born out of frustration with the lack of\nsimple and extendable tooling for mono-repos. It is language-agnostic and\ntoolchain-independent, allowing users to focus on their workflows without being\nconstrained by specific technologies.\n\nQuitsh can be used in two non-coupled ways (enhancing each other):\n\n- Build/use & extend the CLI `quitsh` for your tooling/CI/CD scripts in your\n repository. You write `quitsh mycommand --do-it` and decide what it does,\n using the [library modules](#built-in-libraries) for running subprocesses,\n logging etc.\n\n- Use its [component feature](#component-system) which gives you the ability to\n register your tooling scripts (build/lint/test etc.) as\n [runners](#runner-system) which can be reused across\n [components](#components). Since runners depend heavily on available\n executables in your `PATH`, runners run over a toolchain (currently Nix\n development shell etc.).\n\n## Key Features\n\n### Code-First Approach\n\n- All tooling logic is implemented in `Go`.\n- Tooling logic is defined primarily in code, avoiding declarative\n configurations or templated non-typed languages, which often add unnecessary\n complexity despite their flexibility. _Note: Supporting configuration\n languages like `jsonnet` or \"turing-complete\" YAML etc. is particularly\n avoided. **Everything stays strongly-typed and fast compiled.**_\n\n### Extendability\n\n- `quitsh` serves as a library to build **your customized** CLI tool for your\n specific tasks.\n- Users can add custom commands and specialized tooling features using libraries\n like [`cobra`](https://github.com/spf13/cobra), `quitsh mycommand --do-it`.\n\n### Component System\n\n- [Components](#components) (i.e., buildable units) are identified by placing a\n configuration file (default: `.component.yaml`) in the corresponding\n subdirectory.\n\n#### Targets and Steps\n\n- Each component defines **targets**, which consist of multiple **steps**.\n- Targets can depend on other targets across the repository.\n- Input change sets can be specified for each target to track modifications and\n determine if the target is outdated.\n\n#### Runner System\n\n- Steps within targets are executed by **runners**, which are written by you in\n Go and act as reusable replacements for traditional build/tooling scripts.\n- Runners can have custom YAML configuration options specified per component in\n `.component.yaml`.\n\n#### Toolchain Dispatch\n\n- Runners are associated with specific toolchains.\n- By default, `quitsh` includes a\n [Nix development shell dispatch](https://nix.dev/tutorials/first-steps/declarative-shell.html),\n providing stable and reproducible environments.\n- While container-based dispatching is not a primary goal, it can be implemented\n by extending the dispatch interface.\n\n- The tool was built to replicate the same procedure one executes during local\n development and also in CI. Having CI align with what you execute locally is\n not a nice thing to have, its a necessity. Nix development shells (or\n containers) help with this. A Nix shell provides a simple and robust\n abstraction to pin a toolchain. The following visualization gives an overview\n about how `quitsh` is used:\n\n <p align=\"center\">\n\n \n\n </p>\n\n### Built-in Libraries\n\nThe `pkg` folder offers utilities for common development needs, such as:\n\n- Command Execution: [`pkg/exec`](pkg/exec) provides utilities for process\n execution and command chaining.\n- Structured Logging: [`pkg/log`](pkg/log) enables consistent and readable\n logging.\n- Error Handling: [`pkg/error`](pkg/error) facilitates contextual error\n management.\n- Dependency Graphs: Tools for managing and resolving dependency graphs across\n targets.\n- Some Go `test` runners (here as an example) for running Go tests (its used\n internally to test `quitsh` it-self).\n\n#### Performance\n\n- Since all tooling is written in `Go`, `quitsh` provides type safety and fast\n performance by default.\n- Combined with a Nix-based toolchain dispatch and the ability to write tests\n easily, the framework significantly accelerates the \"change, test, improve\"\n workflow.\n\n#### Nix Integration\n\n- A CLI tool built with `quitsh` can be\n [seamlessly packaged](./tools/nix/packages/cli) into a Nix development shells,\n ensuring accessibility for all users of a given repository.\n\n---\n\n# How To Use It?\n\nUsing this library follows instantiating the CLI (also demonstrated in this\nrepository in [`main.go`](./tools/cli/cmd/cli/main.go), e.g.:\n\n```go\nargs := cliconfig.New()\n\ncli, err := cli.New(\n &args.Commands.Root,\n &args,\n cli.WithName(\"cli\"),\n cli.WithDescription(\"This is the 🐔-🥚 CLI tool for 'quitsh', yes its built with 'quitsh'.\"),\n cli.WithCompFindOptions(\n query.WithFindOptions(\n fs.WithWalkDirFilterPatterns(nil,\n []string{\"**/test/repo/**\"}, true))),\n cli.WithStages(\"lint\", \"build\", \"test\"),\n cli.WithTargetToStageMapperDefault(),\n cli.WithToolchainDispatcherNix(\n \"tools/nix\",\n func(c config.IConfig) *toolchain.DispatchArgs {\n cc := common.Cast[*cliconfig.Config](c)\n\n return &cc.Commands.DispatchArgs\n },\n ),\n)\n```\n\nYou can now add runners and your own commands depending on the needs of your\nrepository. For example in [`main.go`](./tools/cli/cmd/cli/main.go):\n\n```go\nlistcmd.AddCmd(cli, cli.RootCmd())\nconfigcmd.AddCmd(cli.RootCmd(), &conf)\nexectarget.AddCmd(cli, cli.RootCmd())\nexecrunner.AddCmd(cli, cli.RootCmd(), &conf.Commands.DispatchArgs)\n```\n\nadds essential `quitsh` commands\n\n- `listcmd` to list all components etc: `quitsh list`.\n- `configcmd` to inspect/write the config file: `quitsh config ...`.\n- `exectarget` to execute specific targets `quitsh exec-target`.\n- `execrunner` to let `quitsh` dispatch over toolchains (see\n `cli.WithToolchainDispatcherNix` above): `quitsh exec-runner ...`\n\nThere are lots of more useful commands in [`pkg/cli/cmd`](./pkg/cli/cmd) which\nyou might use.\n\n## Useful References\n\nA reference repository with commands and runner can be looked at here:\n\n- [`Custodian`](https://gitlab.com/data-custodian/custodian/tree/main/tools/quitsh)\n- [`DAC-Portal`](https://gitlab.com/data-custodian/dac-portal/tree/main/tools/quitsh)\n- [`Quitsh (this repo)`](https://github.com/sdsc-ordes/quitsh)\n\n## Config\n\nQuitsh runs with global config YAML file which it loads (or defaults) at startup\nfor any invocation. The above CLI instantiation constructs a new config with\n`cliconfig.New()` (this is custom for each usecase and can be adjusted and\nmodified). The config defines global settings (output directories, logging etc.)\nand also various custom, use-case specific settings. These might include\nsettings which runners (or custom commands) might use during execution. For\nexample build runners might use a\n[`build.BuildType`](./tools/cli/pkg/config/config.go) property which could be\n`debug` or `release` etc. The CLI does not care about your custom settings, they\nonly need to be fully serializable to YAML (for toolchain dispatching) and you\ncan override defaults from custom added commands for example.\n\n### Modifying Config Values\n\nYou have the ability to set the config file `quitsh` uses with `--config` or\nread it from stdin with `--config -` or set options (YAML) on the command line\nwith `--config-value`. For example `--config-value \"build.buildType: release\"`\nwould set the `build.BuildType` setting to `release` on startup.\n\n## Components\n\nQuitsh builds around **components**. A **component** should be treated as an own\ncompartment/Git repository or directory with its own independent source code and\noutput.\n\nA component in `quitsh` is defined by a `.component.yaml` (name is\nconfigurable):\n\n```yaml\n# The name of the component: Must be a unique.\nname: my-component\n\n# A semantic version.\n# This is useful for Nix packaging etc.\nversion: 0.2.1\n\n# A simple annotation (not used internally) what main language this component uses.\nlanguage: go\n\n# The `.general` object is not parsed by `quitsh` and\n# allows arbitrary values mainly used for YAML anchors.\n.general:\n value: &val 60\n\ntargets:\n # A target called `test` with two steps.\n my-test:\n # The stage to which this target belongs. Does not need to be provided\n # if the CLI is setup to map target names to stages.\n stage: test\n\n steps:\n # Step 1: Using runner with ID (how it was registered).\n - runner-id: banana-project::my-test-runner\n config: # Your custom runner YAML config, (optional).\n\n # Step 2: Using a runner with registered key (stage: `test`, name `my-test`)\n - runner: my-test\n\n # A target called `build-all` with one step.\n build-all:\n stage: build\n\n # Defining when this target is considered changed:\n # i.e. whenever `self::sources` input change set is changed.\n # `self` maps to this component.\n inputs: [\"self::sources\"]\n\n # Defining dependencies on other targets such that this\n # target is executed after target `my-test` above.\n # You can also link to other components (e.g `other-comp::build`).\n depends: [\"self::my-test\"]\n\n steps:\n # Step 1: Using a runner with registered key (stage: `build`, name `my-test`)\n - runner: my-build\n config:\n tags: [\"doit\"]\n\n lint:\n steps:\n - ... other steps ...\n\ninputs:\n # An input change set with name `sources` which defines\n # patterns to match all source files.\n sources:\n # A regex which matches `*.go` files in `./src` in the components folder.\n patterns:\n - '^./src/.*\\.go$'\n```\n\n## Execution of Targets\n\nThe execution of steps by `quitsh` is done by reading a\n[`.component.yaml`](.component.yaml) for each component. The\n[`.component.yaml`](.component.yaml) file contains _inputs_ and _targets_.\n\nQuitsh's own [`.component.yaml`](./.component.yaml) looks like:\n\n```yaml\nname: quitsh\nlanguage: go\n\ninputs:\n srcs:\n patterns:\n - \"^.*$\"\n\ntargets:\n test:\n stage: test\n steps:\n - runner: go\n - runner: go-bin\n config:\n # Build everything instrumented.\n # Execute the binaries via a `go test` provided in the following\n # pkg and with tags.\n buildPkg: test/cmd\n testPkg: test\n buildTags: [\"integration\"]\n testTags: [\"integration\"]\n\n build:\n steps:\n - runner: go\n lint:\n steps:\n - runner: go\n```\n\nEach **target** defines is a set of **steps** which itself are further specified\nby the field `runner`. A **runner** is **Go code** applicable for a certain step\nwhich should work for all components.\n\nA runner is registered in [`factory`](./pkg/runner/factory/runner.go), for\nexample [here](./tools/cli/pkg/runner/go/lint.go). Runners can be written by\nimplementing the interface [`Runner`](./pkg/runner/runner.go) inside\n[`./pkg/runner/runners`](./pkg/runner/runners) and registering them in\n[`./pkg/runner/factory/init-runners.go`], for example\n[here](./tools/cli/pkg/runner/go/register.go).\n\n> [!NOTE]\n>\n> **You can execute targets in parallel with `--parallel`**.\n\n## Runner Configuration\n\nRunners can load independent YAML config under `config` to make them\nconfigurable, e.g. the `go` build runner loads the following\n[config](./pkg/runner/runners/go/build-config.go):\n\n```yaml\nsteps:\n build:\n - runner: go\n config:\n version-module: \"pkg/myversion-module\" # defaults to `pkg/build`\n```\n\n## Target Stages\n\nEach target also maps to a _stage_ which `quitsh` uses to group targets together\nif you want to find them and make gathering commands such as the example\n[`build` here](https://gitlab.com/data-custodian/custodian/-/blob/main/tools/quitsh/cmd/quitsh/cmd/build/build.go#L83).\nIt collects and runs all targets in the stage `build`.\n\n## How to Extend Functionality\n\nFor users using the `quitsh` CLI framework, it is suggested to follow the\nfollowing points when thinking about new functionality in a repository which\nuses this library. If you need new functionality for CI and local development\nwhich you normally would write in `bash`/`python` follow the following steps:\n\n- If the functionality is **a feature needed in an existing runner and step**:\n Extend the runner and make it work with your new test/build/lint feature.\n\n- If the functionality is **not related to a runner or the same for each\n component with that language**: Extend `quitsh` by providing another\n subcommand which does what you need, see this example\n [`fix-hash`](https://gitlab.com/data-custodian/custodian/-/blob/main/tools/quitsh/cmd/quitsh/cmd/nix/fix-hash/fix.go).\n\n- If the functionality is **for a certain language, e.g. `go` or `python` and\n applies to each component which is written in that language**: consider adding\n a new runner for an already pre-defined stage, e.g. `lint`, `build` etc.\n\n## Example Applications\n\nUnderstand what this framework does, is best accomplished by understanding how\nwe use this framework in our\n[components repo (mono-repo)](https://gitlab.com/data-custodian/custodian). Our\nmajor components are located in\n[./components](https://gitlab.com/data-custodian/custodian/-/tree/main/components).",
|
| 8512 |
"readme_title": "<p align=\"center\">",
|
|
|
|
| 8525 |
"monorepo",
|
| 8526 |
"cli"
|
| 8527 |
],
|
| 8528 |
+
"updated_at": "2026-03-16T08:17:52Z",
|
| 8529 |
"url": "https://github.com/sdsc-ordes/quitsh",
|
| 8530 |
"watchers": 1
|
| 8531 |
}
|
github-account-sdsc-ordes/github_account_github_project_metadata.parquet
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e4786ada217d164c964dd5f0d0ee0c9f517f2329576dd34975a58e64c2e1292d
|
| 3 |
+
size 212405
|
github-account-sdsc-ordes/github_account_github_repo_metadata.json
CHANGED
|
@@ -3736,9 +3736,9 @@
|
|
| 3736 |
"project_foreign_key": null,
|
| 3737 |
"pull_requests_closed": 0,
|
| 3738 |
"pull_requests_merged": 7,
|
| 3739 |
-
"pull_requests_open":
|
| 3740 |
-
"pull_requests_total":
|
| 3741 |
-
"pushed_at": "2026-
|
| 3742 |
"readme_length": 8525,
|
| 3743 |
"readme_text": "<p align=\"center\">\n <img src=\"./docs/assets/mava_logo.svg\" alt=\"project logo\" width=\"250\">\n</p>\n\n<h1 align=\"center\">\n mava-exchange\n</h1>\n<p align=\"center\">\n</p>\n\n\n[](https://github.com/sdsc-ordes/mava-exchange/actions/workflows/normal.yaml)\n[](http://www.apache.org/licenses/LICENSE-2.0.html)\n\n**Authors:**\n\n- [Stefan Milosavljevic](mailto:stefan.milosavljevic@sdsc.ethz.ch)\n- [Sabine Maennel](mailto:sabine.maennel@sdsc.ethz.ch)\n\n# mava-exchange\n\nA Python library and standard package format for exchanging video annotation\ndata between research tools.\n\n## Project Context\n\nThis work is part of the MAVA Project, which aims to improve interoperability\nand data exchange among three research tools — VideoScope, TIB-AV-A, and VIAN.\nBy standardising data formats and developing tools to import, export, and\nvalidate annotation packages, the project enhances data sharing and analysis\ncapabilities across linguistic, media studies, and audiovisual research.\n\nThis infrastructure will enable shared research workflows and ensure adherence\nto FAIR principles, improving the accessibility and reusability of research\ndata.\n\n## The Challenge\n\nVideo analysis via AI is computationally expensive and produces large datasets —\ncontinuous observations such as emotion scores, scene analysis, and audio\nfeatures accumulate quickly across a corpus. It is therefore desirable to share\nthese results among video processing tools without recomputing them.\n\nWhat is needed is a common data exchange format and tools to write, read, and\nvalidate packages in that format. Packages should be:\n\n- Interoperable and self-describing\n- Efficient to write and compact in file size\n\n## mava-exchange\n\n`mava-exchange` addresses this with a standard package format for video\nannotation corpora called `.mediapkg`. The library has two goals:\n\n**Format definition** — `mava-exchange` defines the `.mediapkg` standard: a ZIP\narchive containing annotation data as Parquet files alongside a manifest that\nmaps columns to the MAVA ontology via JSON-LD.\n\n**Tooling** — `mava-exchange` provides a Python library and CLI tools to write,\nread, inspect, and validate `.mediapkg` packages.\n\n## Design Choices\n\n- **ZIP + Parquet** — the `.mediapkg` is a ZIP archive containing one Parquet\n file per annotation track. Parquet offers compact storage, efficient reads,\n and columnar access. ZIP provides a universal container that any tool can open\n for inspection.\n\n- **JSON-LD manifest** — each package contains a `manifest.json` with a JSON-LD\n `@context` that maps Parquet column names to terms in the MAVA ontology. This\n is the semantic layer — it describes what each column means without being part\n of the data itself.\n\n- **MAVA ontology** — the ontology defines a shared vocabulary for annotation\n tracks, time coordinates, and observation dimensions. SHACL shapes are\n included for formal validation.\n\n- **Python** — the library is implemented in Python, as all participating tools\n use Python. Support for other languages may be added in future releases.\n\n## Inspiration\n\nThis format is inspired by GeoJSON and GeoParquet. GeoParquet embeds spatial\nmetadata inside Parquet files to describe geometry columns. `mava-exchange`\napplies the same principle to temporal data: where GeoParquet uses spatial\ncoordinates, `mava-exchange` uses time coordinates on a video timeline. Where\nGeoParquet metadata is purely operational, `mava-exchange` adds a semantic layer\nvia JSON-LD to link columns to a shared ontology.\n\n## Using the library\n\nInstall from PyPI:\n\n```bash\npip install mava-exchange\n# or with uv:\nuv add mava-exchange\n```\n\nWrite, read, and validate `.mediapkg` packages:\n\n```python\nfrom mava_exchange import (\n MediaPackageWriter, MediaPackageReader,\n ObservationSeries, AnnotationSeries, DimensionSpec,\n)\n\n# Define what your tracks mean\nemotions = ObservationSeries(\n name=\"emotions\",\n description=\"Face emotion scores from DeepFace\",\n sampling_interval=0.5,\n dimensions=[\n DimensionSpec(\"angry\", \"Anger probability\", \"[0,1]\"),\n DimensionSpec(\"neutral\", \"Neutral expression\", \"[0,1]\"),\n ]\n)\ntranscript = AnnotationSeries(\n name=\"transcript\",\n description=\"Whisper speech-to-text segments\",\n)\n\n# Write a package\nwith MediaPackageWriter(\"corpus.mediapkg\") as writer:\n writer.add_video(\"video_001\", \"https://example.org/talk.mp4\")\n writer.add_track(\"video_001\", emotions, emotions_df)\n writer.add_track(\"video_001\", transcript, transcript_df)\n\n# Read it back\nwith MediaPackageReader(\"corpus.mediapkg\") as reader:\n df = reader.read_track(\"video_001\", \"emotions\")\n```\n\nTwo CLI tools are also available after installation:\n\n```bash\nmediapkg-inspect corpus.mediapkg\nmediapkg-validate corpus.mediapkg\n```\n\nSee [`docs/tutorial.md`](docs/tutorial.md) for a complete walkthrough.\n\n## Development\n\nClone the repository and install in editable mode with development dependencies:\n\n```bash\ngit clone https://github.com/sdsc-ordes/mava-exchange.git\ncd mava-exchange\nuv sync --group dev\n```\n\nThe project uses [just](https://github.com/casey/just) as a task runner:\n\n```bash\njust test # run the test suite\njust lint # run ruff\njust format # format with ruff and treefmt\njust build # build the package\n```\n\nTo run the example that converts real TSV annotation files into a `.mediapkg`\ncorpus:\n\n```bash\njust example\njust inspect # inspect the resulting corpus.mediapkg\njust validate # validate it\n```\n\n## Further Reading\n\n- [`spec/SPEC.md`](spec/SPEC.md) — full format specification\n- [`spec/mava.ttl`](spec/mava.ttl) — MAVA ontology\n- [`spec/mava.shacl.ttl`](spec/mava.shacl.ttl) — MAVA shacl shapes\n- [`docs/tutorial.md`](docs/tutorial.md) — step-by-step guide\n- `examples`— complete example converting TSV files to `.mediapkg`\n\nFor development:\n\n- [Contribution Guidelines](/CONTRIBUTING.md)\n- [Development Guide](docs/development-guide.md)\n\n## Acknowledgements\n\nThis work was funded by the Swiss Data Science Center (SDSC) through its\nNational Call for Projects as an Infrastructure project.\n\nWe gratefully acknowledge the contributions of the SDSC experts and our\npartners.\n\n**SDSC Experts**:\n\n- Dr. Stefan Milosavljevic, ORCID ID\n [0000-0002-9135-1353](https://orcid.org/0000-0002-9135-1353)\n- Sabine Maennel, ORCID ID\n [0009-0001-3022-8239](https://orcid.org/0009-0001-3022-8239)\n- Robin Franken, ORCID ID\n [0009-0008-0143-9118](https://orcid.org/0009-0008-0143-9118)\n- Dr. Oksana Riba Grognuz, ORCID ID\n [0000-0002-2961-2655](https://orcid.org/0000-0002-2961-2655)\n\n**Partner Institutions**\n\n- Dr. Teodora Vuković, ORCID ID\n [0000-0002-5780-5665](https://orcid.org/0000-0002-5780-5665)\n- Dr. Jeremy Zehr, ORCID ID\n [0000-0002-6046-8647](https://orcid.org/0000-0002-6046-8647)\n- Prof. Dr. Josephine Diecke, ORCID ID\n [0000-0002-9342-0631](https://orcid.org/0000-0002-9342-0631)\n- Dr. Simon Spiegel, ORCID ID\n [0000-0003-2141-5566](https://orcid.org/0000-0003-2141-5566)\n- Prof. Dr. Ralph Ewerth, ORCID ID\n [0000-0003-0918-6297](https://orcid.org/0000-0003-0918-6297)\n- Dr. Eric Müller-Budack, ORCID ID\n [0000-0002-6802-1241](https://orcid.org/0000-0002-6802-1241)\n- Dr. Cristina Grisot, ORCID ID\n [0000-0003-0684-4442](https://orcid.org/0000-0003-0684-4442)\n\n## How to Cite\n\nIf you use this software, please cite it as follows:\n\n👉 See the [CITATION.cff](./CITATION.cff) file for the full list of software\nauthors and citation formats.\n\nWhen referring to the project more broadly (including partner contributions),\nplease acknowledge the funding statement and collaborators listed in the\n[Acknowledgements](#acknowledgements) section:\n\n> \"This work was funded by the Swiss Data Science Center (SDSC) through its\n> National Call for Projects as an Infrastructure project.\"\n\n## Copyright\n\nCopyright © 2025-2026 Swiss Data Science Center\n(SDSC),[www.datascience.ch](http://www.datascience.ch/), ROR:\n[ror.org/02hdt9m26](https://ror.org/02hdt9m26). All rights reserved. The SDSC is\na Swiss National Research Infrastructure, jointly established and legally\nrepresented by the École Polytechnique Fédérale de Lausanne (EPFL) and the\nEidgenössische Technische Hochschule Zürich (ETH Zürich) as a société simple.\nThis copyright encompasses all materials, software, documentation, and other\ncontent created and developed by the SDSC.",
|
| 3744 |
"readme_title": "mava-exchange",
|
|
@@ -6354,7 +6354,7 @@
|
|
| 6354 |
"pull_requests_merged": 2,
|
| 6355 |
"pull_requests_open": 1,
|
| 6356 |
"pull_requests_total": 3,
|
| 6357 |
-
"pushed_at": "2026-03-
|
| 6358 |
"readme_length": 2106,
|
| 6359 |
"readme_text": "<p align=\"center\">\n <img src=\"./docs/assets/logo.svg\" alt=\"project logo\" width=\"250\">\n</p>\n\n<h1 align=\"center\">\n open-pulse-hackathon-analysis\n</h1>\n<p align=\"center\">\n</p>\n\n[](https://github.com/sdsc-ordes/open-pulse-hackathon-analysis/releases/latest)\n[](https://github.com/sdsc-ordes/open-pulse-hackathon-analysis/actions/workflows/normal.yaml)\n\n**Authors:**\n\n- [Eisha Tir Raazia](mailto:eisha.raazia@epfl.ch)\n\n## Installation\n\nDescribe the installation instruction here.\n\n## Usage\n\n### Extract Hackathon Data\n\nExtract data from LauzHack:\n\n```bash\njust hackalysis --output_folder ./data --hackathon_provider lauzhack\n```\n\nExtract specific LauzHack years with the CLI:\n\n```bash\nuv run hackalysis -o ./data -p lauzhack -y 2023,2024,2025\n```\n\n````\nExtract data without specifying the years will default to extracting all available years for the specified provider.\n```bash\nuv run hackalysis -o ./data -p lauzhack\n```\n\nExtract data from Devpost:\n\n```bash\njust hackalysis --output_folder ./data --hackathon_provider devpost --hackathon_name \"ExampleHackathonName\"\n```\nor\n\n```bash\njust hackalysis -o ./data -p devpost -n \"ExampleHackathonName\"\n```\n\n## Development\n\nRead first the [Contribution Guidelines](/CONTRIBUTING.md).\n\nFor technical documentation on setup and development, see the\n[Development Guide](docs/development-guide.md)\n\n## Acknowledgement\n\nAcknowledge all contributors and external collaborators here.\n\n## Copyright\n\nCopyright © 2026-2028 Swiss Data Science Center (SDSC),\n[www.datascience.ch](http://www.datascience.ch/). All rights reserved. The SDSC\nis jointly established and legally represented by the École Polytechnique\nFédérale de Lausanne (EPFL) and the Eidgenössische Technische Hochschule Zürich\n(ETH Zürich). This copyright encompasses all materials, software, documentation,\nand other content created and developed by the SDSC.",
|
| 6360 |
"readme_title": "<p align=\"center\">",
|
|
@@ -7165,11 +7165,11 @@
|
|
| 7165 |
"watchers": 2
|
| 7166 |
},
|
| 7167 |
"https://github.com/sdsc-ordes/quitsh": {
|
| 7168 |
-
"commit_count_default_branch":
|
| 7169 |
"contributors_count": 2,
|
| 7170 |
"contributors_top": [
|
| 7171 |
{
|
| 7172 |
-
"contributions":
|
| 7173 |
"html_url": "https://github.com/gabyx",
|
| 7174 |
"login": "gabyx"
|
| 7175 |
},
|
|
@@ -7311,7 +7311,7 @@
|
|
| 7311 |
"languages_top": [
|
| 7312 |
{
|
| 7313 |
"language": "Go",
|
| 7314 |
-
"size":
|
| 7315 |
},
|
| 7316 |
{
|
| 7317 |
"language": "Nix",
|
|
@@ -7330,8 +7330,8 @@
|
|
| 7330 |
"size": 524
|
| 7331 |
}
|
| 7332 |
],
|
| 7333 |
-
"last_commit_date_default_branch": "2026-03-
|
| 7334 |
-
"last_commit_oid_default_branch": "
|
| 7335 |
"latest_release_date": "2026-03-12T16:51:31Z",
|
| 7336 |
"latest_release_tag": "v0.39.0",
|
| 7337 |
"license_name": "MIT License",
|
|
@@ -7346,7 +7346,7 @@
|
|
| 7346 |
"pull_requests_merged": 59,
|
| 7347 |
"pull_requests_open": 0,
|
| 7348 |
"pull_requests_total": 61,
|
| 7349 |
-
"pushed_at": "2026-03-
|
| 7350 |
"readme_length": 14597,
|
| 7351 |
"readme_text": "<p align=\"center\">\n <img src=\"docs/assets/images/logo.svg\" width=\"400px\">\n</p>\n\n<p align=\"center\">\n <a href=\"https://codecov.io/gh/sdsc-ordes/quitsh\">\n <img src=\"https://codecov.io/gh/sdsc-ordes/quitsh/graph/badge.svg?token=35RP18MHFR\" alt=\"Coverage\" />\n </a>\n <a href=\"https://github.com/sdsc-ordes/quitsh/releases/latest\">\n <img src=\"https://img.shields.io/github/release/sdsc-ordes/quitsh.svg?label=release\" alt=\"Current Release\" />\n </a>\n <a href=\"https://pkg.go.dev/github.com/sdsc-ordes/quitsh\">\n <img src=\"https://pkg.go.dev/badge/github.com/sdsc-ordes/quitsh\" alt=\"Package\" />\n </a>\n <a href=\"https://github.com/sdsc-ordes/quitsh/actions/workflows/normal.yaml\">\n <img src=\"https://img.shields.io/github/actions/workflow/status/sdsc-ordes/quitsh/normal.yaml?label=ci\" alt=\"Pipeline Status\" />\n </a>\n <a href=\"https://mit-license.org/\">\n <img src=\"https://img.shields.io/badge/License-Apache2.0-blue.svg?\" alt=\"License label\" />\n </a>\n</p>\n\n---\n\n> [!CAUTION]\n>\n> This repository is in `beta`. The design space of this tool is still explored\n> and breaking changes might occure. Documentation is incomplete.\n\nThe `quitsh` framework (`/kwɪʧ/`) is a build-tooling **CLI library/framework**\ndesigned to replace loosely-typed scripting languages (e.g., `bash`, `python`,\nand similar alternatives) with the statically-typed language `Go`. Its goal is\nto simplify tooling tasks while providing robust, extendable solutions for\ncomponent repositories (mono-repositories).\n\n`quitsh` is an opinionated framework born out of frustration with the lack of\nsimple and extendable tooling for mono-repos. It is language-agnostic and\ntoolchain-independent, allowing users to focus on their workflows without being\nconstrained by specific technologies.\n\nQuitsh can be used in two non-coupled ways (enhancing each other):\n\n- Build/use & extend the CLI `quitsh` for your tooling/CI/CD scripts in your\n repository. You write `quitsh mycommand --do-it` and decide what it does,\n using the [library modules](#built-in-libraries) for running subprocesses,\n logging etc.\n\n- Use its [component feature](#component-system) which gives you the ability to\n register your tooling scripts (build/lint/test etc.) as\n [runners](#runner-system) which can be reused across\n [components](#components). Since runners depend heavily on available\n executables in your `PATH`, runners run over a toolchain (currently Nix\n development shell etc.).\n\n## Key Features\n\n### Code-First Approach\n\n- All tooling logic is implemented in `Go`.\n- Tooling logic is defined primarily in code, avoiding declarative\n configurations or templated non-typed languages, which often add unnecessary\n complexity despite their flexibility. _Note: Supporting configuration\n languages like `jsonnet` or \"turing-complete\" YAML etc. is particularly\n avoided. **Everything stays strongly-typed and fast compiled.**_\n\n### Extendability\n\n- `quitsh` serves as a library to build **your customized** CLI tool for your\n specific tasks.\n- Users can add custom commands and specialized tooling features using libraries\n like [`cobra`](https://github.com/spf13/cobra), `quitsh mycommand --do-it`.\n\n### Component System\n\n- [Components](#components) (i.e., buildable units) are identified by placing a\n configuration file (default: `.component.yaml`) in the corresponding\n subdirectory.\n\n#### Targets and Steps\n\n- Each component defines **targets**, which consist of multiple **steps**.\n- Targets can depend on other targets across the repository.\n- Input change sets can be specified for each target to track modifications and\n determine if the target is outdated.\n\n#### Runner System\n\n- Steps within targets are executed by **runners**, which are written by you in\n Go and act as reusable replacements for traditional build/tooling scripts.\n- Runners can have custom YAML configuration options specified per component in\n `.component.yaml`.\n\n#### Toolchain Dispatch\n\n- Runners are associated with specific toolchains.\n- By default, `quitsh` includes a\n [Nix development shell dispatch](https://nix.dev/tutorials/first-steps/declarative-shell.html),\n providing stable and reproducible environments.\n- While container-based dispatching is not a primary goal, it can be implemented\n by extending the dispatch interface.\n\n- The tool was built to replicate the same procedure one executes during local\n development and also in CI. Having CI align with what you execute locally is\n not a nice thing to have, its a necessity. Nix development shells (or\n containers) help with this. A Nix shell provides a simple and robust\n abstraction to pin a toolchain. The following visualization gives an overview\n about how `quitsh` is used:\n\n <p align=\"center\">\n\n \n\n </p>\n\n### Built-in Libraries\n\nThe `pkg` folder offers utilities for common development needs, such as:\n\n- Command Execution: [`pkg/exec`](pkg/exec) provides utilities for process\n execution and command chaining.\n- Structured Logging: [`pkg/log`](pkg/log) enables consistent and readable\n logging.\n- Error Handling: [`pkg/error`](pkg/error) facilitates contextual error\n management.\n- Dependency Graphs: Tools for managing and resolving dependency graphs across\n targets.\n- Some Go `test` runners (here as an example) for running Go tests (its used\n internally to test `quitsh` it-self).\n\n#### Performance\n\n- Since all tooling is written in `Go`, `quitsh` provides type safety and fast\n performance by default.\n- Combined with a Nix-based toolchain dispatch and the ability to write tests\n easily, the framework significantly accelerates the \"change, test, improve\"\n workflow.\n\n#### Nix Integration\n\n- A CLI tool built with `quitsh` can be\n [seamlessly packaged](./tools/nix/packages/cli) into a Nix development shells,\n ensuring accessibility for all users of a given repository.\n\n---\n\n# How To Use It?\n\nUsing this library follows instantiating the CLI (also demonstrated in this\nrepository in [`main.go`](./tools/cli/cmd/cli/main.go), e.g.:\n\n```go\nargs := cliconfig.New()\n\ncli, err := cli.New(\n &args.Commands.Root,\n &args,\n cli.WithName(\"cli\"),\n cli.WithDescription(\"This is the 🐔-🥚 CLI tool for 'quitsh', yes its built with 'quitsh'.\"),\n cli.WithCompFindOptions(\n query.WithFindOptions(\n fs.WithWalkDirFilterPatterns(nil,\n []string{\"**/test/repo/**\"}, true))),\n cli.WithStages(\"lint\", \"build\", \"test\"),\n cli.WithTargetToStageMapperDefault(),\n cli.WithToolchainDispatcherNix(\n \"tools/nix\",\n func(c config.IConfig) *toolchain.DispatchArgs {\n cc := common.Cast[*cliconfig.Config](c)\n\n return &cc.Commands.DispatchArgs\n },\n ),\n)\n```\n\nYou can now add runners and your own commands depending on the needs of your\nrepository. For example in [`main.go`](./tools/cli/cmd/cli/main.go):\n\n```go\nlistcmd.AddCmd(cli, cli.RootCmd())\nconfigcmd.AddCmd(cli.RootCmd(), &conf)\nexectarget.AddCmd(cli, cli.RootCmd())\nexecrunner.AddCmd(cli, cli.RootCmd(), &conf.Commands.DispatchArgs)\n```\n\nadds essential `quitsh` commands\n\n- `listcmd` to list all components etc: `quitsh list`.\n- `configcmd` to inspect/write the config file: `quitsh config ...`.\n- `exectarget` to execute specific targets `quitsh exec-target`.\n- `execrunner` to let `quitsh` dispatch over toolchains (see\n `cli.WithToolchainDispatcherNix` above): `quitsh exec-runner ...`\n\nThere are lots of more useful commands in [`pkg/cli/cmd`](./pkg/cli/cmd) which\nyou might use.\n\n## Useful References\n\nA reference repository with commands and runner can be looked at here:\n\n- [`Custodian`](https://gitlab.com/data-custodian/custodian/tree/main/tools/quitsh)\n- [`DAC-Portal`](https://gitlab.com/data-custodian/dac-portal/tree/main/tools/quitsh)\n- [`Quitsh (this repo)`](https://github.com/sdsc-ordes/quitsh)\n\n## Config\n\nQuitsh runs with global config YAML file which it loads (or defaults) at startup\nfor any invocation. The above CLI instantiation constructs a new config with\n`cliconfig.New()` (this is custom for each usecase and can be adjusted and\nmodified). The config defines global settings (output directories, logging etc.)\nand also various custom, use-case specific settings. These might include\nsettings which runners (or custom commands) might use during execution. For\nexample build runners might use a\n[`build.BuildType`](./tools/cli/pkg/config/config.go) property which could be\n`debug` or `release` etc. The CLI does not care about your custom settings, they\nonly need to be fully serializable to YAML (for toolchain dispatching) and you\ncan override defaults from custom added commands for example.\n\n### Modifying Config Values\n\nYou have the ability to set the config file `quitsh` uses with `--config` or\nread it from stdin with `--config -` or set options (YAML) on the command line\nwith `--config-value`. For example `--config-value \"build.buildType: release\"`\nwould set the `build.BuildType` setting to `release` on startup.\n\n## Components\n\nQuitsh builds around **components**. A **component** should be treated as an own\ncompartment/Git repository or directory with its own independent source code and\noutput.\n\nA component in `quitsh` is defined by a `.component.yaml` (name is\nconfigurable):\n\n```yaml\n# The name of the component: Must be a unique.\nname: my-component\n\n# A semantic version.\n# This is useful for Nix packaging etc.\nversion: 0.2.1\n\n# A simple annotation (not used internally) what main language this component uses.\nlanguage: go\n\n# The `.general` object is not parsed by `quitsh` and\n# allows arbitrary values mainly used for YAML anchors.\n.general:\n value: &val 60\n\ntargets:\n # A target called `test` with two steps.\n my-test:\n # The stage to which this target belongs. Does not need to be provided\n # if the CLI is setup to map target names to stages.\n stage: test\n\n steps:\n # Step 1: Using runner with ID (how it was registered).\n - runner-id: banana-project::my-test-runner\n config: # Your custom runner YAML config, (optional).\n\n # Step 2: Using a runner with registered key (stage: `test`, name `my-test`)\n - runner: my-test\n\n # A target called `build-all` with one step.\n build-all:\n stage: build\n\n # Defining when this target is considered changed:\n # i.e. whenever `self::sources` input change set is changed.\n # `self` maps to this component.\n inputs: [\"self::sources\"]\n\n # Defining dependencies on other targets such that this\n # target is executed after target `my-test` above.\n # You can also link to other components (e.g `other-comp::build`).\n depends: [\"self::my-test\"]\n\n steps:\n # Step 1: Using a runner with registered key (stage: `build`, name `my-test`)\n - runner: my-build\n config:\n tags: [\"doit\"]\n\n lint:\n steps:\n - ... other steps ...\n\ninputs:\n # An input change set with name `sources` which defines\n # patterns to match all source files.\n sources:\n # A regex which matches `*.go` files in `./src` in the components folder.\n patterns:\n - '^./src/.*\\.go$'\n```\n\n## Execution of Targets\n\nThe execution of steps by `quitsh` is done by reading a\n[`.component.yaml`](.component.yaml) for each component. The\n[`.component.yaml`](.component.yaml) file contains _inputs_ and _targets_.\n\nQuitsh's own [`.component.yaml`](./.component.yaml) looks like:\n\n```yaml\nname: quitsh\nlanguage: go\n\ninputs:\n srcs:\n patterns:\n - \"^.*$\"\n\ntargets:\n test:\n stage: test\n steps:\n - runner: go\n - runner: go-bin\n config:\n # Build everything instrumented.\n # Execute the binaries via a `go test` provided in the following\n # pkg and with tags.\n buildPkg: test/cmd\n testPkg: test\n buildTags: [\"integration\"]\n testTags: [\"integration\"]\n\n build:\n steps:\n - runner: go\n lint:\n steps:\n - runner: go\n```\n\nEach **target** defines is a set of **steps** which itself are further specified\nby the field `runner`. A **runner** is **Go code** applicable for a certain step\nwhich should work for all components.\n\nA runner is registered in [`factory`](./pkg/runner/factory/runner.go), for\nexample [here](./tools/cli/pkg/runner/go/lint.go). Runners can be written by\nimplementing the interface [`Runner`](./pkg/runner/runner.go) inside\n[`./pkg/runner/runners`](./pkg/runner/runners) and registering them in\n[`./pkg/runner/factory/init-runners.go`], for example\n[here](./tools/cli/pkg/runner/go/register.go).\n\n> [!NOTE]\n>\n> **You can execute targets in parallel with `--parallel`**.\n\n## Runner Configuration\n\nRunners can load independent YAML config under `config` to make them\nconfigurable, e.g. the `go` build runner loads the following\n[config](./pkg/runner/runners/go/build-config.go):\n\n```yaml\nsteps:\n build:\n - runner: go\n config:\n version-module: \"pkg/myversion-module\" # defaults to `pkg/build`\n```\n\n## Target Stages\n\nEach target also maps to a _stage_ which `quitsh` uses to group targets together\nif you want to find them and make gathering commands such as the example\n[`build` here](https://gitlab.com/data-custodian/custodian/-/blob/main/tools/quitsh/cmd/quitsh/cmd/build/build.go#L83).\nIt collects and runs all targets in the stage `build`.\n\n## How to Extend Functionality\n\nFor users using the `quitsh` CLI framework, it is suggested to follow the\nfollowing points when thinking about new functionality in a repository which\nuses this library. If you need new functionality for CI and local development\nwhich you normally would write in `bash`/`python` follow the following steps:\n\n- If the functionality is **a feature needed in an existing runner and step**:\n Extend the runner and make it work with your new test/build/lint feature.\n\n- If the functionality is **not related to a runner or the same for each\n component with that language**: Extend `quitsh` by providing another\n subcommand which does what you need, see this example\n [`fix-hash`](https://gitlab.com/data-custodian/custodian/-/blob/main/tools/quitsh/cmd/quitsh/cmd/nix/fix-hash/fix.go).\n\n- If the functionality is **for a certain language, e.g. `go` or `python` and\n applies to each component which is written in that language**: consider adding\n a new runner for an already pre-defined stage, e.g. `lint`, `build` etc.\n\n## Example Applications\n\nUnderstand what this framework does, is best accomplished by understanding how\nwe use this framework in our\n[components repo (mono-repo)](https://gitlab.com/data-custodian/custodian). Our\nmajor components are located in\n[./components](https://gitlab.com/data-custodian/custodian/-/tree/main/components).",
|
| 7352 |
"readme_title": "<p align=\"center\">",
|
|
@@ -7365,7 +7365,7 @@
|
|
| 7365 |
"monorepo",
|
| 7366 |
"cli"
|
| 7367 |
],
|
| 7368 |
-
"updated_at": "2026-03-
|
| 7369 |
"url": "https://github.com/sdsc-ordes/quitsh",
|
| 7370 |
"watchers": 1
|
| 7371 |
},
|
|
|
|
| 3736 |
"project_foreign_key": null,
|
| 3737 |
"pull_requests_closed": 0,
|
| 3738 |
"pull_requests_merged": 7,
|
| 3739 |
+
"pull_requests_open": 2,
|
| 3740 |
+
"pull_requests_total": 9,
|
| 3741 |
+
"pushed_at": "2026-03-16T13:25:15Z",
|
| 3742 |
"readme_length": 8525,
|
| 3743 |
"readme_text": "<p align=\"center\">\n <img src=\"./docs/assets/mava_logo.svg\" alt=\"project logo\" width=\"250\">\n</p>\n\n<h1 align=\"center\">\n mava-exchange\n</h1>\n<p align=\"center\">\n</p>\n\n\n[](https://github.com/sdsc-ordes/mava-exchange/actions/workflows/normal.yaml)\n[](http://www.apache.org/licenses/LICENSE-2.0.html)\n\n**Authors:**\n\n- [Stefan Milosavljevic](mailto:stefan.milosavljevic@sdsc.ethz.ch)\n- [Sabine Maennel](mailto:sabine.maennel@sdsc.ethz.ch)\n\n# mava-exchange\n\nA Python library and standard package format for exchanging video annotation\ndata between research tools.\n\n## Project Context\n\nThis work is part of the MAVA Project, which aims to improve interoperability\nand data exchange among three research tools — VideoScope, TIB-AV-A, and VIAN.\nBy standardising data formats and developing tools to import, export, and\nvalidate annotation packages, the project enhances data sharing and analysis\ncapabilities across linguistic, media studies, and audiovisual research.\n\nThis infrastructure will enable shared research workflows and ensure adherence\nto FAIR principles, improving the accessibility and reusability of research\ndata.\n\n## The Challenge\n\nVideo analysis via AI is computationally expensive and produces large datasets —\ncontinuous observations such as emotion scores, scene analysis, and audio\nfeatures accumulate quickly across a corpus. It is therefore desirable to share\nthese results among video processing tools without recomputing them.\n\nWhat is needed is a common data exchange format and tools to write, read, and\nvalidate packages in that format. Packages should be:\n\n- Interoperable and self-describing\n- Efficient to write and compact in file size\n\n## mava-exchange\n\n`mava-exchange` addresses this with a standard package format for video\nannotation corpora called `.mediapkg`. The library has two goals:\n\n**Format definition** — `mava-exchange` defines the `.mediapkg` standard: a ZIP\narchive containing annotation data as Parquet files alongside a manifest that\nmaps columns to the MAVA ontology via JSON-LD.\n\n**Tooling** — `mava-exchange` provides a Python library and CLI tools to write,\nread, inspect, and validate `.mediapkg` packages.\n\n## Design Choices\n\n- **ZIP + Parquet** — the `.mediapkg` is a ZIP archive containing one Parquet\n file per annotation track. Parquet offers compact storage, efficient reads,\n and columnar access. ZIP provides a universal container that any tool can open\n for inspection.\n\n- **JSON-LD manifest** — each package contains a `manifest.json` with a JSON-LD\n `@context` that maps Parquet column names to terms in the MAVA ontology. This\n is the semantic layer — it describes what each column means without being part\n of the data itself.\n\n- **MAVA ontology** — the ontology defines a shared vocabulary for annotation\n tracks, time coordinates, and observation dimensions. SHACL shapes are\n included for formal validation.\n\n- **Python** — the library is implemented in Python, as all participating tools\n use Python. Support for other languages may be added in future releases.\n\n## Inspiration\n\nThis format is inspired by GeoJSON and GeoParquet. GeoParquet embeds spatial\nmetadata inside Parquet files to describe geometry columns. `mava-exchange`\napplies the same principle to temporal data: where GeoParquet uses spatial\ncoordinates, `mava-exchange` uses time coordinates on a video timeline. Where\nGeoParquet metadata is purely operational, `mava-exchange` adds a semantic layer\nvia JSON-LD to link columns to a shared ontology.\n\n## Using the library\n\nInstall from PyPI:\n\n```bash\npip install mava-exchange\n# or with uv:\nuv add mava-exchange\n```\n\nWrite, read, and validate `.mediapkg` packages:\n\n```python\nfrom mava_exchange import (\n MediaPackageWriter, MediaPackageReader,\n ObservationSeries, AnnotationSeries, DimensionSpec,\n)\n\n# Define what your tracks mean\nemotions = ObservationSeries(\n name=\"emotions\",\n description=\"Face emotion scores from DeepFace\",\n sampling_interval=0.5,\n dimensions=[\n DimensionSpec(\"angry\", \"Anger probability\", \"[0,1]\"),\n DimensionSpec(\"neutral\", \"Neutral expression\", \"[0,1]\"),\n ]\n)\ntranscript = AnnotationSeries(\n name=\"transcript\",\n description=\"Whisper speech-to-text segments\",\n)\n\n# Write a package\nwith MediaPackageWriter(\"corpus.mediapkg\") as writer:\n writer.add_video(\"video_001\", \"https://example.org/talk.mp4\")\n writer.add_track(\"video_001\", emotions, emotions_df)\n writer.add_track(\"video_001\", transcript, transcript_df)\n\n# Read it back\nwith MediaPackageReader(\"corpus.mediapkg\") as reader:\n df = reader.read_track(\"video_001\", \"emotions\")\n```\n\nTwo CLI tools are also available after installation:\n\n```bash\nmediapkg-inspect corpus.mediapkg\nmediapkg-validate corpus.mediapkg\n```\n\nSee [`docs/tutorial.md`](docs/tutorial.md) for a complete walkthrough.\n\n## Development\n\nClone the repository and install in editable mode with development dependencies:\n\n```bash\ngit clone https://github.com/sdsc-ordes/mava-exchange.git\ncd mava-exchange\nuv sync --group dev\n```\n\nThe project uses [just](https://github.com/casey/just) as a task runner:\n\n```bash\njust test # run the test suite\njust lint # run ruff\njust format # format with ruff and treefmt\njust build # build the package\n```\n\nTo run the example that converts real TSV annotation files into a `.mediapkg`\ncorpus:\n\n```bash\njust example\njust inspect # inspect the resulting corpus.mediapkg\njust validate # validate it\n```\n\n## Further Reading\n\n- [`spec/SPEC.md`](spec/SPEC.md) — full format specification\n- [`spec/mava.ttl`](spec/mava.ttl) — MAVA ontology\n- [`spec/mava.shacl.ttl`](spec/mava.shacl.ttl) — MAVA shacl shapes\n- [`docs/tutorial.md`](docs/tutorial.md) — step-by-step guide\n- `examples`— complete example converting TSV files to `.mediapkg`\n\nFor development:\n\n- [Contribution Guidelines](/CONTRIBUTING.md)\n- [Development Guide](docs/development-guide.md)\n\n## Acknowledgements\n\nThis work was funded by the Swiss Data Science Center (SDSC) through its\nNational Call for Projects as an Infrastructure project.\n\nWe gratefully acknowledge the contributions of the SDSC experts and our\npartners.\n\n**SDSC Experts**:\n\n- Dr. Stefan Milosavljevic, ORCID ID\n [0000-0002-9135-1353](https://orcid.org/0000-0002-9135-1353)\n- Sabine Maennel, ORCID ID\n [0009-0001-3022-8239](https://orcid.org/0009-0001-3022-8239)\n- Robin Franken, ORCID ID\n [0009-0008-0143-9118](https://orcid.org/0009-0008-0143-9118)\n- Dr. Oksana Riba Grognuz, ORCID ID\n [0000-0002-2961-2655](https://orcid.org/0000-0002-2961-2655)\n\n**Partner Institutions**\n\n- Dr. Teodora Vuković, ORCID ID\n [0000-0002-5780-5665](https://orcid.org/0000-0002-5780-5665)\n- Dr. Jeremy Zehr, ORCID ID\n [0000-0002-6046-8647](https://orcid.org/0000-0002-6046-8647)\n- Prof. Dr. Josephine Diecke, ORCID ID\n [0000-0002-9342-0631](https://orcid.org/0000-0002-9342-0631)\n- Dr. Simon Spiegel, ORCID ID\n [0000-0003-2141-5566](https://orcid.org/0000-0003-2141-5566)\n- Prof. Dr. Ralph Ewerth, ORCID ID\n [0000-0003-0918-6297](https://orcid.org/0000-0003-0918-6297)\n- Dr. Eric Müller-Budack, ORCID ID\n [0000-0002-6802-1241](https://orcid.org/0000-0002-6802-1241)\n- Dr. Cristina Grisot, ORCID ID\n [0000-0003-0684-4442](https://orcid.org/0000-0003-0684-4442)\n\n## How to Cite\n\nIf you use this software, please cite it as follows:\n\n👉 See the [CITATION.cff](./CITATION.cff) file for the full list of software\nauthors and citation formats.\n\nWhen referring to the project more broadly (including partner contributions),\nplease acknowledge the funding statement and collaborators listed in the\n[Acknowledgements](#acknowledgements) section:\n\n> \"This work was funded by the Swiss Data Science Center (SDSC) through its\n> National Call for Projects as an Infrastructure project.\"\n\n## Copyright\n\nCopyright © 2025-2026 Swiss Data Science Center\n(SDSC),[www.datascience.ch](http://www.datascience.ch/), ROR:\n[ror.org/02hdt9m26](https://ror.org/02hdt9m26). All rights reserved. The SDSC is\na Swiss National Research Infrastructure, jointly established and legally\nrepresented by the École Polytechnique Fédérale de Lausanne (EPFL) and the\nEidgenössische Technische Hochschule Zürich (ETH Zürich) as a société simple.\nThis copyright encompasses all materials, software, documentation, and other\ncontent created and developed by the SDSC.",
|
| 3744 |
"readme_title": "mava-exchange",
|
|
|
|
| 6354 |
"pull_requests_merged": 2,
|
| 6355 |
"pull_requests_open": 1,
|
| 6356 |
"pull_requests_total": 3,
|
| 6357 |
+
"pushed_at": "2026-03-16T09:09:15Z",
|
| 6358 |
"readme_length": 2106,
|
| 6359 |
"readme_text": "<p align=\"center\">\n <img src=\"./docs/assets/logo.svg\" alt=\"project logo\" width=\"250\">\n</p>\n\n<h1 align=\"center\">\n open-pulse-hackathon-analysis\n</h1>\n<p align=\"center\">\n</p>\n\n[](https://github.com/sdsc-ordes/open-pulse-hackathon-analysis/releases/latest)\n[](https://github.com/sdsc-ordes/open-pulse-hackathon-analysis/actions/workflows/normal.yaml)\n\n**Authors:**\n\n- [Eisha Tir Raazia](mailto:eisha.raazia@epfl.ch)\n\n## Installation\n\nDescribe the installation instruction here.\n\n## Usage\n\n### Extract Hackathon Data\n\nExtract data from LauzHack:\n\n```bash\njust hackalysis --output_folder ./data --hackathon_provider lauzhack\n```\n\nExtract specific LauzHack years with the CLI:\n\n```bash\nuv run hackalysis -o ./data -p lauzhack -y 2023,2024,2025\n```\n\n````\nExtract data without specifying the years will default to extracting all available years for the specified provider.\n```bash\nuv run hackalysis -o ./data -p lauzhack\n```\n\nExtract data from Devpost:\n\n```bash\njust hackalysis --output_folder ./data --hackathon_provider devpost --hackathon_name \"ExampleHackathonName\"\n```\nor\n\n```bash\njust hackalysis -o ./data -p devpost -n \"ExampleHackathonName\"\n```\n\n## Development\n\nRead first the [Contribution Guidelines](/CONTRIBUTING.md).\n\nFor technical documentation on setup and development, see the\n[Development Guide](docs/development-guide.md)\n\n## Acknowledgement\n\nAcknowledge all contributors and external collaborators here.\n\n## Copyright\n\nCopyright © 2026-2028 Swiss Data Science Center (SDSC),\n[www.datascience.ch](http://www.datascience.ch/). All rights reserved. The SDSC\nis jointly established and legally represented by the École Polytechnique\nFédérale de Lausanne (EPFL) and the Eidgenössische Technische Hochschule Zürich\n(ETH Zürich). This copyright encompasses all materials, software, documentation,\nand other content created and developed by the SDSC.",
|
| 6360 |
"readme_title": "<p align=\"center\">",
|
|
|
|
| 7165 |
"watchers": 2
|
| 7166 |
},
|
| 7167 |
"https://github.com/sdsc-ordes/quitsh": {
|
| 7168 |
+
"commit_count_default_branch": 217,
|
| 7169 |
"contributors_count": 2,
|
| 7170 |
"contributors_top": [
|
| 7171 |
{
|
| 7172 |
+
"contributions": 216,
|
| 7173 |
"html_url": "https://github.com/gabyx",
|
| 7174 |
"login": "gabyx"
|
| 7175 |
},
|
|
|
|
| 7311 |
"languages_top": [
|
| 7312 |
{
|
| 7313 |
"language": "Go",
|
| 7314 |
+
"size": 414373
|
| 7315 |
},
|
| 7316 |
{
|
| 7317 |
"language": "Nix",
|
|
|
|
| 7330 |
"size": 524
|
| 7331 |
}
|
| 7332 |
],
|
| 7333 |
+
"last_commit_date_default_branch": "2026-03-16T08:16:51Z",
|
| 7334 |
+
"last_commit_oid_default_branch": "fcb9abac7c966f92ffe91dc496dc7506020d1280",
|
| 7335 |
"latest_release_date": "2026-03-12T16:51:31Z",
|
| 7336 |
"latest_release_tag": "v0.39.0",
|
| 7337 |
"license_name": "MIT License",
|
|
|
|
| 7346 |
"pull_requests_merged": 59,
|
| 7347 |
"pull_requests_open": 0,
|
| 7348 |
"pull_requests_total": 61,
|
| 7349 |
+
"pushed_at": "2026-03-16T08:17:48Z",
|
| 7350 |
"readme_length": 14597,
|
| 7351 |
"readme_text": "<p align=\"center\">\n <img src=\"docs/assets/images/logo.svg\" width=\"400px\">\n</p>\n\n<p align=\"center\">\n <a href=\"https://codecov.io/gh/sdsc-ordes/quitsh\">\n <img src=\"https://codecov.io/gh/sdsc-ordes/quitsh/graph/badge.svg?token=35RP18MHFR\" alt=\"Coverage\" />\n </a>\n <a href=\"https://github.com/sdsc-ordes/quitsh/releases/latest\">\n <img src=\"https://img.shields.io/github/release/sdsc-ordes/quitsh.svg?label=release\" alt=\"Current Release\" />\n </a>\n <a href=\"https://pkg.go.dev/github.com/sdsc-ordes/quitsh\">\n <img src=\"https://pkg.go.dev/badge/github.com/sdsc-ordes/quitsh\" alt=\"Package\" />\n </a>\n <a href=\"https://github.com/sdsc-ordes/quitsh/actions/workflows/normal.yaml\">\n <img src=\"https://img.shields.io/github/actions/workflow/status/sdsc-ordes/quitsh/normal.yaml?label=ci\" alt=\"Pipeline Status\" />\n </a>\n <a href=\"https://mit-license.org/\">\n <img src=\"https://img.shields.io/badge/License-Apache2.0-blue.svg?\" alt=\"License label\" />\n </a>\n</p>\n\n---\n\n> [!CAUTION]\n>\n> This repository is in `beta`. The design space of this tool is still explored\n> and breaking changes might occure. Documentation is incomplete.\n\nThe `quitsh` framework (`/kwɪʧ/`) is a build-tooling **CLI library/framework**\ndesigned to replace loosely-typed scripting languages (e.g., `bash`, `python`,\nand similar alternatives) with the statically-typed language `Go`. Its goal is\nto simplify tooling tasks while providing robust, extendable solutions for\ncomponent repositories (mono-repositories).\n\n`quitsh` is an opinionated framework born out of frustration with the lack of\nsimple and extendable tooling for mono-repos. It is language-agnostic and\ntoolchain-independent, allowing users to focus on their workflows without being\nconstrained by specific technologies.\n\nQuitsh can be used in two non-coupled ways (enhancing each other):\n\n- Build/use & extend the CLI `quitsh` for your tooling/CI/CD scripts in your\n repository. You write `quitsh mycommand --do-it` and decide what it does,\n using the [library modules](#built-in-libraries) for running subprocesses,\n logging etc.\n\n- Use its [component feature](#component-system) which gives you the ability to\n register your tooling scripts (build/lint/test etc.) as\n [runners](#runner-system) which can be reused across\n [components](#components). Since runners depend heavily on available\n executables in your `PATH`, runners run over a toolchain (currently Nix\n development shell etc.).\n\n## Key Features\n\n### Code-First Approach\n\n- All tooling logic is implemented in `Go`.\n- Tooling logic is defined primarily in code, avoiding declarative\n configurations or templated non-typed languages, which often add unnecessary\n complexity despite their flexibility. _Note: Supporting configuration\n languages like `jsonnet` or \"turing-complete\" YAML etc. is particularly\n avoided. **Everything stays strongly-typed and fast compiled.**_\n\n### Extendability\n\n- `quitsh` serves as a library to build **your customized** CLI tool for your\n specific tasks.\n- Users can add custom commands and specialized tooling features using libraries\n like [`cobra`](https://github.com/spf13/cobra), `quitsh mycommand --do-it`.\n\n### Component System\n\n- [Components](#components) (i.e., buildable units) are identified by placing a\n configuration file (default: `.component.yaml`) in the corresponding\n subdirectory.\n\n#### Targets and Steps\n\n- Each component defines **targets**, which consist of multiple **steps**.\n- Targets can depend on other targets across the repository.\n- Input change sets can be specified for each target to track modifications and\n determine if the target is outdated.\n\n#### Runner System\n\n- Steps within targets are executed by **runners**, which are written by you in\n Go and act as reusable replacements for traditional build/tooling scripts.\n- Runners can have custom YAML configuration options specified per component in\n `.component.yaml`.\n\n#### Toolchain Dispatch\n\n- Runners are associated with specific toolchains.\n- By default, `quitsh` includes a\n [Nix development shell dispatch](https://nix.dev/tutorials/first-steps/declarative-shell.html),\n providing stable and reproducible environments.\n- While container-based dispatching is not a primary goal, it can be implemented\n by extending the dispatch interface.\n\n- The tool was built to replicate the same procedure one executes during local\n development and also in CI. Having CI align with what you execute locally is\n not a nice thing to have, its a necessity. Nix development shells (or\n containers) help with this. A Nix shell provides a simple and robust\n abstraction to pin a toolchain. The following visualization gives an overview\n about how `quitsh` is used:\n\n <p align=\"center\">\n\n \n\n </p>\n\n### Built-in Libraries\n\nThe `pkg` folder offers utilities for common development needs, such as:\n\n- Command Execution: [`pkg/exec`](pkg/exec) provides utilities for process\n execution and command chaining.\n- Structured Logging: [`pkg/log`](pkg/log) enables consistent and readable\n logging.\n- Error Handling: [`pkg/error`](pkg/error) facilitates contextual error\n management.\n- Dependency Graphs: Tools for managing and resolving dependency graphs across\n targets.\n- Some Go `test` runners (here as an example) for running Go tests (its used\n internally to test `quitsh` it-self).\n\n#### Performance\n\n- Since all tooling is written in `Go`, `quitsh` provides type safety and fast\n performance by default.\n- Combined with a Nix-based toolchain dispatch and the ability to write tests\n easily, the framework significantly accelerates the \"change, test, improve\"\n workflow.\n\n#### Nix Integration\n\n- A CLI tool built with `quitsh` can be\n [seamlessly packaged](./tools/nix/packages/cli) into a Nix development shells,\n ensuring accessibility for all users of a given repository.\n\n---\n\n# How To Use It?\n\nUsing this library follows instantiating the CLI (also demonstrated in this\nrepository in [`main.go`](./tools/cli/cmd/cli/main.go), e.g.:\n\n```go\nargs := cliconfig.New()\n\ncli, err := cli.New(\n &args.Commands.Root,\n &args,\n cli.WithName(\"cli\"),\n cli.WithDescription(\"This is the 🐔-🥚 CLI tool for 'quitsh', yes its built with 'quitsh'.\"),\n cli.WithCompFindOptions(\n query.WithFindOptions(\n fs.WithWalkDirFilterPatterns(nil,\n []string{\"**/test/repo/**\"}, true))),\n cli.WithStages(\"lint\", \"build\", \"test\"),\n cli.WithTargetToStageMapperDefault(),\n cli.WithToolchainDispatcherNix(\n \"tools/nix\",\n func(c config.IConfig) *toolchain.DispatchArgs {\n cc := common.Cast[*cliconfig.Config](c)\n\n return &cc.Commands.DispatchArgs\n },\n ),\n)\n```\n\nYou can now add runners and your own commands depending on the needs of your\nrepository. For example in [`main.go`](./tools/cli/cmd/cli/main.go):\n\n```go\nlistcmd.AddCmd(cli, cli.RootCmd())\nconfigcmd.AddCmd(cli.RootCmd(), &conf)\nexectarget.AddCmd(cli, cli.RootCmd())\nexecrunner.AddCmd(cli, cli.RootCmd(), &conf.Commands.DispatchArgs)\n```\n\nadds essential `quitsh` commands\n\n- `listcmd` to list all components etc: `quitsh list`.\n- `configcmd` to inspect/write the config file: `quitsh config ...`.\n- `exectarget` to execute specific targets `quitsh exec-target`.\n- `execrunner` to let `quitsh` dispatch over toolchains (see\n `cli.WithToolchainDispatcherNix` above): `quitsh exec-runner ...`\n\nThere are lots of more useful commands in [`pkg/cli/cmd`](./pkg/cli/cmd) which\nyou might use.\n\n## Useful References\n\nA reference repository with commands and runner can be looked at here:\n\n- [`Custodian`](https://gitlab.com/data-custodian/custodian/tree/main/tools/quitsh)\n- [`DAC-Portal`](https://gitlab.com/data-custodian/dac-portal/tree/main/tools/quitsh)\n- [`Quitsh (this repo)`](https://github.com/sdsc-ordes/quitsh)\n\n## Config\n\nQuitsh runs with global config YAML file which it loads (or defaults) at startup\nfor any invocation. The above CLI instantiation constructs a new config with\n`cliconfig.New()` (this is custom for each usecase and can be adjusted and\nmodified). The config defines global settings (output directories, logging etc.)\nand also various custom, use-case specific settings. These might include\nsettings which runners (or custom commands) might use during execution. For\nexample build runners might use a\n[`build.BuildType`](./tools/cli/pkg/config/config.go) property which could be\n`debug` or `release` etc. The CLI does not care about your custom settings, they\nonly need to be fully serializable to YAML (for toolchain dispatching) and you\ncan override defaults from custom added commands for example.\n\n### Modifying Config Values\n\nYou have the ability to set the config file `quitsh` uses with `--config` or\nread it from stdin with `--config -` or set options (YAML) on the command line\nwith `--config-value`. For example `--config-value \"build.buildType: release\"`\nwould set the `build.BuildType` setting to `release` on startup.\n\n## Components\n\nQuitsh builds around **components**. A **component** should be treated as an own\ncompartment/Git repository or directory with its own independent source code and\noutput.\n\nA component in `quitsh` is defined by a `.component.yaml` (name is\nconfigurable):\n\n```yaml\n# The name of the component: Must be a unique.\nname: my-component\n\n# A semantic version.\n# This is useful for Nix packaging etc.\nversion: 0.2.1\n\n# A simple annotation (not used internally) what main language this component uses.\nlanguage: go\n\n# The `.general` object is not parsed by `quitsh` and\n# allows arbitrary values mainly used for YAML anchors.\n.general:\n value: &val 60\n\ntargets:\n # A target called `test` with two steps.\n my-test:\n # The stage to which this target belongs. Does not need to be provided\n # if the CLI is setup to map target names to stages.\n stage: test\n\n steps:\n # Step 1: Using runner with ID (how it was registered).\n - runner-id: banana-project::my-test-runner\n config: # Your custom runner YAML config, (optional).\n\n # Step 2: Using a runner with registered key (stage: `test`, name `my-test`)\n - runner: my-test\n\n # A target called `build-all` with one step.\n build-all:\n stage: build\n\n # Defining when this target is considered changed:\n # i.e. whenever `self::sources` input change set is changed.\n # `self` maps to this component.\n inputs: [\"self::sources\"]\n\n # Defining dependencies on other targets such that this\n # target is executed after target `my-test` above.\n # You can also link to other components (e.g `other-comp::build`).\n depends: [\"self::my-test\"]\n\n steps:\n # Step 1: Using a runner with registered key (stage: `build`, name `my-test`)\n - runner: my-build\n config:\n tags: [\"doit\"]\n\n lint:\n steps:\n - ... other steps ...\n\ninputs:\n # An input change set with name `sources` which defines\n # patterns to match all source files.\n sources:\n # A regex which matches `*.go` files in `./src` in the components folder.\n patterns:\n - '^./src/.*\\.go$'\n```\n\n## Execution of Targets\n\nThe execution of steps by `quitsh` is done by reading a\n[`.component.yaml`](.component.yaml) for each component. The\n[`.component.yaml`](.component.yaml) file contains _inputs_ and _targets_.\n\nQuitsh's own [`.component.yaml`](./.component.yaml) looks like:\n\n```yaml\nname: quitsh\nlanguage: go\n\ninputs:\n srcs:\n patterns:\n - \"^.*$\"\n\ntargets:\n test:\n stage: test\n steps:\n - runner: go\n - runner: go-bin\n config:\n # Build everything instrumented.\n # Execute the binaries via a `go test` provided in the following\n # pkg and with tags.\n buildPkg: test/cmd\n testPkg: test\n buildTags: [\"integration\"]\n testTags: [\"integration\"]\n\n build:\n steps:\n - runner: go\n lint:\n steps:\n - runner: go\n```\n\nEach **target** defines is a set of **steps** which itself are further specified\nby the field `runner`. A **runner** is **Go code** applicable for a certain step\nwhich should work for all components.\n\nA runner is registered in [`factory`](./pkg/runner/factory/runner.go), for\nexample [here](./tools/cli/pkg/runner/go/lint.go). Runners can be written by\nimplementing the interface [`Runner`](./pkg/runner/runner.go) inside\n[`./pkg/runner/runners`](./pkg/runner/runners) and registering them in\n[`./pkg/runner/factory/init-runners.go`], for example\n[here](./tools/cli/pkg/runner/go/register.go).\n\n> [!NOTE]\n>\n> **You can execute targets in parallel with `--parallel`**.\n\n## Runner Configuration\n\nRunners can load independent YAML config under `config` to make them\nconfigurable, e.g. the `go` build runner loads the following\n[config](./pkg/runner/runners/go/build-config.go):\n\n```yaml\nsteps:\n build:\n - runner: go\n config:\n version-module: \"pkg/myversion-module\" # defaults to `pkg/build`\n```\n\n## Target Stages\n\nEach target also maps to a _stage_ which `quitsh` uses to group targets together\nif you want to find them and make gathering commands such as the example\n[`build` here](https://gitlab.com/data-custodian/custodian/-/blob/main/tools/quitsh/cmd/quitsh/cmd/build/build.go#L83).\nIt collects and runs all targets in the stage `build`.\n\n## How to Extend Functionality\n\nFor users using the `quitsh` CLI framework, it is suggested to follow the\nfollowing points when thinking about new functionality in a repository which\nuses this library. If you need new functionality for CI and local development\nwhich you normally would write in `bash`/`python` follow the following steps:\n\n- If the functionality is **a feature needed in an existing runner and step**:\n Extend the runner and make it work with your new test/build/lint feature.\n\n- If the functionality is **not related to a runner or the same for each\n component with that language**: Extend `quitsh` by providing another\n subcommand which does what you need, see this example\n [`fix-hash`](https://gitlab.com/data-custodian/custodian/-/blob/main/tools/quitsh/cmd/quitsh/cmd/nix/fix-hash/fix.go).\n\n- If the functionality is **for a certain language, e.g. `go` or `python` and\n applies to each component which is written in that language**: consider adding\n a new runner for an already pre-defined stage, e.g. `lint`, `build` etc.\n\n## Example Applications\n\nUnderstand what this framework does, is best accomplished by understanding how\nwe use this framework in our\n[components repo (mono-repo)](https://gitlab.com/data-custodian/custodian). Our\nmajor components are located in\n[./components](https://gitlab.com/data-custodian/custodian/-/tree/main/components).",
|
| 7352 |
"readme_title": "<p align=\"center\">",
|
|
|
|
| 7365 |
"monorepo",
|
| 7366 |
"cli"
|
| 7367 |
],
|
| 7368 |
+
"updated_at": "2026-03-16T08:17:52Z",
|
| 7369 |
"url": "https://github.com/sdsc-ordes/quitsh",
|
| 7370 |
"watchers": 1
|
| 7371 |
},
|
github-account-sdsc-ordes/github_account_github_repo_metadata.parquet
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b45bb3c8fb9ef0d026e33e6e1be43f45cec327a47e2a76c999806dd072367251
|
| 3 |
+
size 206684
|