WIP: publishing metadata document schema

schema/deploy.json

@@ -0,0 +1,130 @@
+{
+  // URL to the schema definition, which we should make public.
+  "$schema": "./posit-publishing-schema-v3.json",
+
+  // Defines what we are deploying
+  "type": "quarto-static",
+  "entrypoint": "report.qmd",
+
+  // These metadata fields are purely for human consumption.
+  "title": "Regional Quarterly Sales Report",
+  "description": "This is the quarterly sales report, broken down by region.",
+  "thumbnail": "images/thumbnail.jpg",
+  "author": {
+    "name": "Michael Marchetti",
+    "email": "[email protected]",
+    "username": "mmarchetti",
+    "organization": "Posit, PBC"
+  },
+  // Arbitrary string tags. On Connect, only admins can create tags,
+  // and tags can be nested (e.g. "sales,reports,quarterly" is a
+  // single tag 3 levels deep in the tag tree). The example below
+  // is 3 separate top level tags.
+  "tags": [
+    "sales",
+    "quarterly",
+    "regional"
+  ],
+  // List of (relative) file paths. If constructing a bundle to upload,
+  // these files must be included. For git deployments, the server
+  // will expect (required) that these files be present in the repo.
+  // Unlike manifest v1, there are no checksums, since those require
+  // updating this file whenever other files change, and we'd prefer
+  // to write this once and commit it.
+  "files": [
+    "report.qmd",
+    "_quarto.yml",
+    "model.py",
+    "requirements.txt",
+    "weights/sales-model.hdf5"
+  ],
+  "dependencies": {
+    "python": {
+      "version": "3.11.3",
+      "package_file": "requirements.txt",
+      "package_manager": "pip"
+    },
+    "r": {
+      "version": "4.3.1",
+      "package_file": "renv.lock",
+      "package_manager": "renv"
+    },
+    "quarto": {
+      "version": "1.4"
+    }
+  },
+  "environment": {
+      "API_URL": "https://example.com/api",
+      // secret value, will be pulled from the environment when deploying through CLI,
+      // must be set at the server if deploying through git.
+      "API_KEY": null
+  },
+  "schedule": {
+    // Only valid for reports, not for apps.
+    // iCalendar spec: https://icalendar.org/iCalendar-RFC-5545/3-8-5-3-recurrence-rule.html
+    "start": "2023-10-25T08:00:00Z",
+    "recurrence": "FREQ=MONTHLY;INTERVAL=3"
+  },
+  "access": {
+    "users": [
+      {
+        // Password auth uses usernames as unique IDs.
+        // You can specify the auth provider unique ID,
+        // Connect-assigned GUID, or both. Tools
+        // should populate both, because GUIDs are not
+        // portable across Connect instances.
+        "id": "jqpublic",
+        "guid": "536b456e-0311-4f92-ba10-dbf1db8a468e",
+        "name": "John Q. Public",
+        "permissions": "editor"
+      }
+    ],
+    "groups": [
+      {
+        // Connect identifies groups with a name and a GUID,
+        // both of which are unique. LDAP may provide
+        // a CN/DN as the unique ID, with a separate name. Tools
+        // should populate both, because GUIDs are not
+        // portable across Connect instances.
+        "id": "Data Science Team",
+        "guid": "8b4fde3e-f995-4894-bc02-ae47538262ff",
+        "name": "Data Science Team",
+        "permissions": "editor"
+      }
+    ]
+  },
+  "connect": {
+    // These settings are Connect-specific and will be ignored by other servers.
+    "access": {
+      "run_as": "rstudio-connect",
+      "run_as_current_user": false
+    },
+    "runtime": {
+      // Might some of these apply beyond Connect? For example, init_timeout
+      // tells the server how long this app might take to start up.
+      "connection_timeout": 5,
+      "read_timeout": 30,
+      "init_timeout": 60,
+      "idle_timeout": 120,
+      "max_processes": 5,
+      "min_processes": 1,
+      "max_conns_per_process": 50,
+      "load_factor": 0.5
+    },
+    "kubernetes": {
+      // Will Posit Cloud Connect be containerized? If so, will it support
+      // image_name, cpu_limit, memory_limit, and/or gpu settings?
+      // (note k8s GPU settings are different from the docker ones)
+      "amd_gpu_limit": 0,
+      "cpu_limit": 1,
+      "cpu_request": 0.5,
+      "image_name": "posit/connect-runtime-python3.11-r4.3",
+      "memory_limit": "100000000",
+      "memory_request": "20000000",
+      "nvidia_gpu_limit": 0,
+      "service_account_name": "posit-connect-content",
+      "r_environment_management": true,
+      "py_environment_management": true
+    }
+  }
+}

schema/deploy.toml

@@ -0,0 +1,78 @@
+"$schema" = "./posit-publishing-schema-v3.json"
+type = "quarto-static"
+entrypoint = "report.qmd"
+title = "Regional Quarterly Sales Report"
+description = "This is the quarterly sales report, broken down by region."
+thumbnail = "images/thumbnail.jpg"
+tags = [ "sales", "quarterly", "regional" ]
+files = [
+  "report.qmd",
+  "_quarto.yml",
+  "model.py",
+  "requirements.txt",
+  "weights/sales-model.hdf5"
+]
+
+[author]
+name = "Michael Marchetti"
+email = "[email protected]"
+username = "mmarchetti"
+organization = "Posit, PBC"
+
+[dependencies.python]
+version = "3.11.3"
+package_file = "requirements.txt"
+package_manager = "pip"
+
+[dependencies.r]
+version = "4.3.1"
+package_file = "renv.lock"
+package_manager = "renv"
+
+[dependencies.quarto]
+version = "1.4"
+
+[environment]
+API_URL = "https://example.com/api"
+
+[schedule]
+start = "2023-10-25T08:00:00Z"
+recurrence = "FREQ=MONTHLY;INTERVAL=3"
+
+[[access.users]]
+id = "jqpublic"
+guid = "536b456e-0311-4f92-ba10-dbf1db8a468e"
+name = "John Q. Public"
+permissions = "editor"
+
+[[access.groups]]
+id = "Data Science Team"
+guid = "8b4fde3e-f995-4894-bc02-ae47538262ff"
+name = "Data Science Team"
+permissions = "editor"
+
+[connect.access]
+run_as = "rstudio-connect"
+run_as_current_user = false
+
+[connect.runtime]
+connection_timeout = 5
+read_timeout = 30
+init_timeout = 60
+idle_timeout = 120
+max_processes = 5
+min_processes = 1
+max_conns_per_process = 50
+load_factor = 0.5
+
+[connect.kubernetes]
+amd_gpu_limit = 0
+cpu_limit = 1
+cpu_request = 0.5
+image_name = "posit/connect-runtime-python3.11-r4.3"
+memory_limit = "100000000"
+memory_request = "20000000"
+nvidia_gpu_limit = 0
+service_account_name = "posit-connect-content"
+r_environment_management = true
+py_environment_management = true

schema/deploy.yml

@@ -0,0 +1,74 @@
+# yaml-language-server: $schema=./posit-publishing-schema-v3.json
+"$schema": "./posit-publishing-schema-v3.json"
+type: quarto-static
+entrypoint: report.qmd
+title: Regional Quarterly Sales Report
+description: This is the quarterly sales report, broken down by region.
+thumbnail: images/thumbnail.jpg
+author:
+  name: Michael Marchetti
+  email: [email protected]
+  username: mmarchetti
+  organization: Posit, PBC
+tags:
+- sales
+- quarterly
+- regional
+files:
+- report.qmd
+- _quarto.yml
+- model.py
+- requirements.txt
+- weights/sales-model.hdf5
+dependencies:
+  python:
+    version: 3.11.3
+    package_file: requirements.txt
+    package_manager: pip
+  r:
+    version: 4.3.1
+    package_file: renv.lock
+    package_manager: renv
+  quarto:
+    version: '1.4'
+environment:
+  API_URL: https://example.com/api
+  API_KEY:
+schedule:
+  start: '2023-10-25T08:00:00Z'
+  recurrence: FREQ=MONTHLY;INTERVAL=3
+access:
+  users:
+  - id: jqpublic
+    guid: 536b456e-0311-4f92-ba10-dbf1db8a468e
+    name: John Q. Public
+    permissions: editor
+  groups:
+  - id: Data Science Team
+    guid: 8b4fde3e-f995-4894-bc02-ae47538262ff
+    name: Data Science Team
+    permissions: editor
+connect:
+  access:
+    run_as: rstudio-connect
+    run_as_current_user: false
+  runtime:
+    connection_timeout: 5
+    read_timeout: 30
+    init_timeout: 60
+    idle_timeout: 120
+    max_processes: 5
+    min_processes: 1
+    max_conns_per_process: 50
+    load_factor: 0.5
+  kubernetes:
+    amd_gpu_limit: 0
+    cpu_limit: 1
+    cpu_request: 0.5
+    image_name: posit/connect-runtime-python3.11-r4.3
+    memory_limit: '100000000'
+    memory_request: '20000000'
+    nvidia_gpu_limit: 0
+    service_account_name: posit-connect-content
+    r_environment_management: true
+    py_environment_management: true

schema/posit-publishing-record-schema-v3.json

@@ -0,0 +1,48 @@
+{
+  "$schema": "https://json-schema.org/draft/2020-12/schema",
+  "$id": "https://github.com/rstudio/publishing-client/posit-publishing-record-schema-v3.json",
+  "type": "object",
+  "additionalProperties": false,
+  "description": "Posit Publishing Record",
+  "required": [
+    "$schema",
+    "server-url",
+    "id",
+    "configuration"
+  ],
+  "properties": {
+    "$schema": {
+      "type": "string",
+      "format": "url",
+      "description": "URL of the json-schema definition for this file. Must be 'https://cdn.posit.co/connect/posit-publishing-schema-v3.json'.",
+      "enum": [
+        "./posit-publishing-record-schema-v3.json"
+      ],
+      "examples": [
+        "./posit-publishing-record-schema-v3.json"
+      ]
+    },
+    "server-url": {
+      "type": "string",
+      "format": "uri",
+      "description": "URL of the server where this content was deployed.",
+      "examples": [
+        "https://connect.example.com"
+      ]
+    },
+    "id": {
+      "type": "string",
+      "description": "Unique ID of this deployment.",
+      "examples": [
+        "de2e7bdb-b085-401e-a65c-443e40009749"
+      ]
+    },
+    "configuration": {
+      "type": "string",
+      "description": "Path to the configuration file within the project directory.",
+      "examples": [
+        "production.json"
+      ]
+    }
+  }
+}