Changes for page Technical details

Last modified by lzehl on 2021/07/05 18:57

From 40.1 to 39.1 From 80.1 to 79.1

From version 79.1

edited by lzehl
on 2021/07/05 10:34

Change comment: There is no comment for this version

To version 40.1

edited by lzehl
on 2021/03/09 16:52

Change comment: There is no comment for this version

Raw
Rendered

Summary

Page properties (2 modified, 0 added, 0 removed)

Details

Page properties

Parent

@@ -1,1 +1,1 @@
--Collabs.openminds.Documentation.WebHome
++Collabs.openminds.openMINDS core.WebHome

Content

@@ -10,155 +10,150 @@
  Please find below a documentation of the layout and requirements needed to keep the openMINDS modularity, the syntax of the openMINDS schema template, as well as the openMINDS integration pipeline.
  )))
--=== The openMINDS umbrella ===
++=== Overview of the openMINDS layout ===
  (% style="text-align: justify;" %)
--openMINDS is the overall umbrella for a set of integrated metadata models for describing neuroscience research products in graph databases. The correct integration of these metadata models as well as the provision of a central access point is handled by an integration pipeline. All metadata models are developed on separate GitHub repositories, in order to facilitate extensions and maintenance of existing, as well as development and integration of new openMINDS metadata models and schemas. The integration pipeline ensures that the central openMINDS GitHub repository ingests all these GitHub repositories as [[git-submodules>>https://git-scm.com/docs/git-submodule||rel="noopener noreferrer" target="_blank"]], integrates the respective metadata models and builds the openMINDS GitHub pages, as well as ZIP files containing the respective openMINDS schemas in the supported formats, such as the openMINDS syntax (cf. below), JSON-Schema, or HTML. In the following we will summarize the contents and requirements for the central openMINDS GitHub repository as well as all metadata model git-submodules. The openMINDS integration pipeline will be handled in a separate chapter (cf. below).
++In summary, openMINDS is the overall umbrella for a set of distributed GitHub repositories, each defining a particular metadata model for neuroscience research products.
--(% style="text-align: justify;" %)
--Let us start with the [[central openMINDS GitHub repository>>https://github.com/HumanBrainProject/openMINDS||rel="noopener noreferrer" target="_blank"]] which has a **main branch**, a **documentation branch**, and **version branches** (naming convention: vX; e.g., v1). Official releases (naming convention: vX.Y; e.g., v1.0) are tagged and provided as release packages.
++=== The openMINDS schema template syntax ===
  (% style="text-align: justify;" %)
--The **main branch** hosts the general README, the LICENSE document, the CONTRIBUTING document, and the general openMINDS logo. In addition, it maintains the openMINDS vocabulary (vocab; cf. below) which provides general definitions and references for schema types and properties used across all openMINDS metadata models and their versions, as well as the bash script that builds the content of the documentation and version branches.
++All openMINDS metadata models use a light-weighted schema template syntax for defining the expected metadata. The correspondingly formatted schema files use the extension: **##.schema.tpl.json##**.
  (% style="text-align: justify;" %)
--The **documentation** branch hosts the HTML files that build the [[openMINDS GitHub pages>>https://humanbrainproject.github.io/openMINDS/||rel="noopener noreferrer" target="_blank"]], as well as a ZIP file for each version branch and official release containing the respective openMINDS schemas in the currently supported formats, such as the openMINDS syntax (`.schema.tpl.json`; cf. below), JSON-Schema (`.schema.json`), or HTML (`.html`).
++Although, as the file extension suggests, this openMINDS schema template syntax is inspired by JSON-Schema, it facilitates or even excludes technical aspects that are generally expected for the openMINDS schemas making them more human-readable, especially for untrained eyes. Behind the scenes, within the openMINDS integration pipeline (cf. below), this schema template syntax is then interpreted and flexibly translated to various formal metadata formats (e.g., JSON-Schema).
  (% style="text-align: justify;" %)
--The **version** branches host the respective openMINDS schemas of a major version by ingesting the corresponding metadata models as git-submodules. Note that these version branches can have official release tags. If a version branch has an official release tag, only backwards compatible changes can be merged on this branch. This can include corrections of typos in instructions, introduction of additional properties to schemas, loosening constraints on expected value numbers or formats, granting additional relations between schemas, and adding new schemas (if they do not require relational changes in existing schemas). Except for typo corrections, these changes are typically tagged as sub-releases for the respective major version (e.g., v1.1). If a version branch does not have an official release tag, yet, also non-backwards compatible changes can be merged on this branch. This can include renaming of existing properties, increasing constraints on expected value numbers or formats, removing relations between schemas and adding new schemas, if they cause relational changes in existing schemas. In case all version branches have official release tags, a new non-backwards compatible change would lead to the creation of a new version branch (with a respectively increased major version number).
++Despite the simplification in comparison to JSON-Schema, the openMINDS schema templates are also, at the core, specially formatted JSON files using a particular syntax, meaning special key-value pairs that define the validation rules of a schema.
  (% style="text-align: justify;" %)
--As mentioned already above, the setup of the central openMINDS GitHub repository is maintained by the openMINDS integration pipeline (cf. below). Note that the pipeline is configured in such a way, that each commit on one of the openMINDS submodules will trigger a new build of the central openMINDS repository ensuring that its content is always up-to-date.
++Please find in the following a full documentation of the openMINDS schema template syntax and how it's key-value pairs need to be defined and interpreted.
  (% style="text-align: justify;" %)
--For this to work smoothly for the existing, but also for all new openMINDS metadata models, the corresponding openMINDS submodules (GitHub repositories) have to meet the following requirements:
++===== Target & concept templates =====
  (% style="text-align: justify;" %)
--**(1)** The openMINDS metadata model has to be located on a **public GitHub repository** and published under an **MIT license**.
++Same as in JSON-Schema, all openMINDS schema templates define the expected name (written in **##lowerCamelCase##**) and value of the metadata, typically called property, under the key **##properties##** as nested dictionaries. Furthermore, the names of obligatory metadata can be listed under the key **##required##**. Here a generalized example:
--(% style="text-align: justify;" %)
--**(2)** The GitHub repository of such a metadata model should have at least one **version branch** (naming convention: vX, where X is a major version number).
++{{code language="json"}}
++{
++ "properties": {
++   "propertyNameA": {},
++   "propertyNameB": {},
++   "propertyNameC": {}
++ },
++ "required": [
++   "propertyNameA",
++   "propertyNameC"
++ ]
++}
++{{/code}}
  (% style="text-align: justify;" %)
--**(3)** The version branch should have the following **folders** & **files**:
++In addition, an openMINDS schema //has to have// a key **##"_type"##** to be recognized as **target template**. In other words, the **##"_type"##** is used to define the openMINDS namespace of a corresponding schema using a particular naming convention. Here again a generalized example:
--* **##schemas/##** (required) - contains the schemas of the respective metadata model implemented in the reduced openMINDS syntax (cf. below). The sub-directory of the "schemas" folder can be further structured or flat.
--* **##tests/##** (recommended) - contains test-instances (JSON-LDs) for checking the constraints defined in the schemas of the respective metadata model. The sub-directory of the "tests" folder should follow the same structure as the "schemas" folder with an additional sub-directory for each schema. The file names of the test-instances should be written in lowerCamelCase and state first the name of the respectively tested schema and second, separated with an underscore, what schema constrain is tested (e.g., **##contactInformation_validEmail.jsonld##**). If a test-instance is expected to fail the schema validation the file name should receive the postfix **##"_nok"##** (e.g., **##contactInformation_invalidEmail_nok.jsonld##**).
--* **##examples/##** (recommended) - contains examples for valid instance collections for the respective metadata model. Each example should receive its own directory (folder) with a **##README.md##** describing the example, and an **##metadataCollection##** subfolder containing the openMINDS instances (JSON-LDs). This subfolder can be further structured or flat.
--* **##img/##** (optional) - contains typically the logo of the openMINDS submodule.
--* **##instances/##** (optional) - contains the controlled metadata instances (JSON-LDs) for selected schemas of the respective metadata model. The sub-directory of the "instances" folder should follow the same structure as the "schemas" folder with an additional sub-directory for each schema. The file names of the controlled instances should be written in lowerCamelCase and state the simple, human-readable identifier of the instance (e.g., **##homoSapiens.jsonld##** for the controlledTerms schema **##Species##**).
--* **##version.txt##** (required) - states the identifier of the version branch (e.g., **##v1##**) of the respective metadata model.
--* **##README.md##** (required) - contains a short content description of the respective metadata model.
--* **##LICENSE.txt##** (required) - states the license for the respective metadata model (has to be MIT).
--
--=== The openMINDS vocabulary ===
--
--(% style="text-align: justify;" %)
--Located under the folder **##vocab##** in the main openMINDS GitHub directory, the openMINDS vocabulary is semi-automatically gathered and stored in dedicated JSON files ([[**##types.json##**>>https://raw.githubusercontent.com/HumanBrainProject/openMINDS/v2/vocab/types.json||rel="noopener noreferrer" target="_blank"]] and [[**##properties.json##**>>https://raw.githubusercontent.com/HumanBrainProject/openMINDS/v2/vocab/properties.json||rel="noopener noreferrer" target="_blank"]]). The openMINDS integration pipeline makes sure that both files are updated with each commit to any of the GitHub repositories for the openMINDS metadata models. With that, the openMINDS vocab reflects always an up-to-date status of the general attributes of existing **schemas** and **properties** across all openMINDS metadata models, while providing the opportunity to centrally review and maintain their consistency. In addition, this design allows us to centrally define and maintain multiple references to related schemas and matching schema properties of other metadata initiatives. How this works in detail is explained in the following.
--
--(% style="text-align: justify;" %)
--The **##types.json##** file is an associative array listing all existing openMINDS schemas (via their type). For each openMINDS schema, a small list of general attributes are provided in a nested associative array. Currently, the following attributes are captured:
--
  {{code language="json"}}
  {
--  "OPENMINDS_SCHEMA_TYPE": {
--    "description": "GENERAL_DESCRIPTION",
--    "name": "DISPLAY_LABEL",
--    "translatableTo": [
--      "REFERENCE_TO_RELATED_SCHEMA_OF_OTHER_INITIATIVE"
--    ]
--  }
++ "_type": "https:~/~/openminds.ebrains.eu/<<schema-model>>/<<schema-name>>",
++ "properties": {}
  }
  {{/code}}
  (% style="text-align: justify;" %)
--With each new schema committed to one of the openMINDS metadata models, a new entry is appended to the **##types.json##** file, with the display label automatically derived from the respective schema type and the remaining attributes predefined with a null value. Once an entry for a schema is made in the **##types.json##** file, the values of all attributes (**##"name"##**, **##"description"##**, and **##"translatableTo"##**) can be manually edited. All manual editions will be preserved and not overwritten when the file is updated again with a new commit. In case a schema is deleted from the openMINDS metadata models, the corresponding entry in the **##types.json##** file is marked as being deprecated (additional attribute-value pair; **##"deprecated": true##**). It only can be permanently removed from the **##types.json##** file, if the entry is manually deleted.
++Note that **##<<schema-model>>##** has to be replaced with the label of the openMINDS metadata model to which the corresponding schema belongs to, and **##<<schema-name>>##** has to be replaced with the corresponding name of the schema (written in **##CamelCase##**).
  (% style="text-align: justify;" %)
--Similar to the **##types.json##** file, the **##properties.json##** file is an associative array listing all properties across all existing openMINDS schemas (via the property name). For each openMINDS property, a small list of general attributes are provided in a nested associative array. Currently, the following attributes are captured:
++If an openMINDS schema template //does not// define a key **##"_type"##** (as in the first example above), it is interpreted as a **concept template** which //has to be// extended to a target template.
++(% style="text-align: justify;" %)
++Concept templates are and should be used when multiple target templates have the same subset of properties, because they facilitate the long-term maintenance of those shared properties: Instead of defining the same properties repeatedly within multiple target templates, the common subset can be defined within a single concept template and passed on to all extending target templates.
++
++(% style="text-align: justify;" %)
++To define that a target template is the extension of a concept template, the target template can state under **##"_extends"##** the relative path to the concept template. For example, the openMINDS core target template **##Dataset##** extends the core concept template **##researchProduct##** as indicated here:
++
  {{code language="json"}}
  {
--  "PROPERTY_NAME": {
--    "description": "GENERAL_DESCRIPTION",
--    "name": "DISPLAY_LABEL",
--    "nameForReverseLink": "DISPLAY_LABEL_OF_REVERSED_LINK",
--    "sameAs": [
--      "REFERENCE_TO_MATCHING_SCHEMA-PROPERTY_OF_OTHER_INITIATIVE"
--    ],
--    "schemas": [
--      "RELATIVE_PATH_TO_OPENMINDS-SCHEMA_USING_THIS_PROPERTY"
--    ]
--  }
++  "_type": "https:~/~/openminds.ebrains.eu/core/Dataset",
++  "_extends": "products/researchProduct.schema.tpl.json"
  }
  {{/code}}
  (% style="text-align: justify;" %)
--With each new property committed to a schema of one of the openMINDS metadata models, a new entry is appended to the **##properties.json##** file, with the display label and list of schemas in which this property occurs automatically derived. The remaining attributes are initially provided with a null value. Once an entry for a property is made in the **##properties.json##** file, the values of all attributes (**##"name"##**, **##"description"##**, **##"nameForReversedLink"##**, and **##"sameAs"##**) can be manually edited, except for **##"schemas"##** which will be always automatically updated. All those manual editions will be preserved and not overwritten when the file is updated again with a new commit. In case a property is not used anymore in any of the schemas from the openMINDS metadata models, the corresponding entry in the **##properties.json##** file is marked as being deprecated (additional attribute-value pair; **##"deprecated": true##**). It only can be permanently removed from the **##properties.json##** file, if the entry is manually deleted.
++Note that this convention requires the concept and corresponding target templates to be located in the same openMINDS metadata model repository. Note also that for properties, the following rules apply for target and concept template:
--=== The openMINDS syntax ===
++1. A concept template has to define some properties which will be inherited by all extending target templates.
++1. If a concept template additionally states that some of these properties are required, all extending target templates will require the same properties.
++1. A target template can require properties of the concept template, that are not explicitly required within the concept template. In such a case, the other target templates extending the same concept template will not require those properties.
++1. A target template can (but does not have to) define and require additional properties that were not defined and required in the concept template. These additionally defined and required properties will not be shared with the other target templates extending the same concept template.
  (% style="text-align: justify;" %)
--All openMINDS metadata models are defined using a light-weighted schema syntax. Although this schema syntax is inspired by JSON-Schema, it outsources most schema technicalities to be handled by the openMINDS integration pipeline, making the openMINDS schemas more human-readable, especially for untrained eyes.
++How to define the expected value of a property will be explained for the different property types in the following sections.
  (% style="text-align: justify;" %)
--The few remaining customized technical properties which need additional interpretation or translation to a formal schema languages (e.g. JSON-Schema) have an underscore as prefix (e.g., **##"_type"##**). Within the openMINDS integration pipeline (cf. below), the schema template syntax is interpreted, extended and flexibly translated to various formal schema languages. All further specifications of the openMINDS schema template syntax are described below.
++===== Properties expecting a string value =====
  (% style="text-align: justify;" %)
--==== Basic openMINDS schema structure ====
++Properties that expect values of type string can be defined in large parts in the same way as in JSON-Schema. Without any further formatting requirements the definition of such a property looks like this:
--(% style="text-align: justify;" %)
--All openMINDS schemas need to have the extension **##.schema.tpl.json##** and each schema is defined as a nested associative array (dictionary) with the following conceptual structure:
--
  {{code language="json"}}
  {
--  "_type": "https://openminds.ebrains.eu/LABEL_OF_METADATA_MODEL/SCHEMA_NAME",
    "properties": {
--    "PROPERTY_NAME": {
--      "type": "DATA_TYPE",
--      "_instruction": "METADATA_ENTRY_INSTRUCTION"
--  },
--  "required": [
--    "PROPERTY_NAME"
--  ]
++    "freeStringProperty": {
++      "type": "string"
++    }
++  }
  }
  {{/code}}
  (% style="text-align: justify;" %)
--**##"_type"##** defines the schema type (or namespace) with the depicted naming convention, where the label of the respective openMINDS metadata model (e.g., **##"core"##**) and the schema name (format: UpperCamelCase; e.g. **##"ContactInformation"##**) have to be specified. Obviously, the schema name should be meaningful and provide some insides into what metadata content the schema covers.
++In accordance with JSON-Schema, it is possible, though, to further define a selected list of built-in formats that are accepted for the expected string. Supported are all [[built-in formats of JSON-Schema Draft 7.0>>https://json-schema.org/understanding-json-schema/reference/string.html#built-in-formats||rel="noopener noreferrer" target="_blank"]]. Assuming that my string property only accepts a value of format "email", the openMINDS syntax is the following:
--(% style="text-align: justify;" %)
--Under **##"properties"##** a nested associative array is defined, where each key defines the property name (format: lowerCamelCase; e.g. **##"givenName"##**). The corresponding value is again a nested associative array defining the expected data **##"type"##** (cf. below) and the **##"_instructions"##** for entering the correct metadata for the respective property.
++{{code language="json"}}
++{
++  "properties": {
++    "emailProperty": {
++      "type": "string",
++      "_formats": [
++        "email"
++      ]
++    }
++  }
++}
++{{/code}}
  (% style="text-align: justify;" %)
--Under **##"required"##** a list of property names can be provided that are obligatory to be present in a correctly instantiated metadata instance of the respective schema. If none of the properties are required, this key-value pair does not have to be specified.
++Also like JSON-Schema, it is possible to define a customized regular expression pattern for the expected string, supporting the [[ECMA-262  regex-dialect>>https://www.ecma-international.org/publications-and-standards/standards/ecma-262/]]. Assuming that my string property only accepts a Gregorian calendar year, the corresponding openMINDS syntax look like this:
--(% style="text-align: justify;" %)
--==== Schemas extending a context-schema ====
++{{code language="json"}}
++{
++  "properties": {
++    "yearProperty": {
++      "type": "string",
++      "pattern": "([0-9]{4})"
++    }
++  }
++}
++{{/code}}
  (% style="text-align: justify;" %)
--In the case that several schemas are highly related and contain a common set of properties, it is possible to define a non-type context-schema with these common properties that can be extended and modified by the group of related schemas.
++===== Properties expecting a numerical value =====
  (% style="text-align: justify;" %)
--All properties and constraints (e.g. required properties, expected data types) defined in the context-schema are passed on to the schemas extending this context-schema. Each of these schemas can define additional properties, or (if necessary) can overwrite the constraints of the context-schema (incl. **##"_instructions"##**). In order to state that a schema is extending a context-schema, the following additional key-value pair has to be added to the schema template above:
++(//**coming soon**//)
--(% style="text-align: center;" %)
--**##"_extends": "RELATIVE_PATH_TO_OPENMINDS-CONTEXT-SCHEMA"##**
--
  (% style="text-align: justify;" %)
--This design not only makes it easier to identify highly related schemas, but also facilitates the maintenance of the commonly used properties. A good hands-on example, is the context-schema [[ResearchProduct>>https://raw.githubusercontent.com/HumanBrainProject/openMINDS_core/v3/schemas/products/researchProduct.schema.tpl.json||rel="noopener noreferrer" target="_blank"]] which is extended by the following schema set: [[Dataset>>https://raw.githubusercontent.com/HumanBrainProject/openMINDS_core/v3/schemas/products/dataset.schema.tpl.json||rel="noopener noreferrer" target="_blank"]], [[MetaDataModel>>https://raw.githubusercontent.com/HumanBrainProject/openMINDS_core/v3/schemas/products/metaDataModel.schema.tpl.json]], [[Model>>https://raw.githubusercontent.com/HumanBrainProject/openMINDS_core/v3/schemas/products/model.schema.tpl.json||rel="noopener noreferrer" target="_blank"]], and [[Software>>https://raw.githubusercontent.com/HumanBrainProject/openMINDS_core/v3/schemas/products/software.schema.tpl.json||rel="noopener noreferrer" target="_blank"]].
++===== Properties expecting another object =====
  (% style="text-align: justify;" %)
--==== Data type depending constraints ====
++(//**coming soon**//)
  (% style="text-align: justify;" %)
--Depending on the expected data **##"type"##** additional constraints can be made for the metadata entry of a respective property. Currently, the openMINDS schema template syntax supports the following data types: **##"string"##**, ##**"integer"**##, **##"float"##**, **##"boolean"##**, **##"array"##** and **##"object"##**.
++===== Properties expecting an array of values =====
  (% style="text-align: justify;" %)
--If the expected data **##"type"##** is a **##"string"##** the expected number of characters, the format or a regular expression pattern of the string can be further defined.
++(//**coming soon**//)
++
  === The openMINDS integration pipeline ===
--(//**coming soon**//) If you'd like to learn more about the openMINDS integration pipeline, especially if you'd like to contribute to it, please get in touch with us (the openMINDS development team) via the issues on the openMINDS or openMINDS_generator GitHub or the support email: openminds@ebrains.eu
--
--{{putFootnotes/}}
++(//**coming soon**//)