Scripts

process_repo

Usage: process_repo [--logfile logfile] [--output output] [--db file|mongo|github] [--imports true|false] dir1 dir2 ...

This is the main script used to process vendor repositories and update the central FMU data. The dir* arguments indicates the directories to process.

Options

• --db file|github - Indicates where data should be read from and written to. (default: file)
• --repo ... - The GitHub repository to read from and write data to. Note, this option only applies when the --db is set to github. Note that pushing data to GitHub is only supported if the environment variable GITHUB_TOKEN is supplied. Otherwise, the push step will be skipped and the only real work done by the script will be validation. (default: [email protected]:xogeny/fmi-standard.org.git)
• --branch ... - Specifies which branch should be used when --db is set to github. (default: master)
• --output ... - The directory to read from and write data to. Note, this only applies when the --db is set to file. If no value for output is specified, then no legacy data will be used to initialize the processing and no data will be written (see use cases for more details).
• --imports true|false - If set to true, import (i.e., corss-check) data is ignored. (default: false)
• --pedantic true|false - If set to true, scripts are stricter in what they accept. (default: false)
• --moved true|false - If set to true, tools are not checked against previous vendor ids. This should only be turned on if a tool is suddenly owned by a different vendor. (default: false).
• --logfile ... - Name of file to write output to. If nothing is specified, output will be written to standard out.

Use Cases

To simple test processing of a directory defaults values should be sufficient for basic tests, e.g.,

$ process_repo <DIR>

In this case, no legacy data is read in and none of the data created will be written. This is how vendors can perform some basic validation on their data.

To save the resulting data, simply specify an output directory, e.g.,

$ process_repo <DIR> --output <OUTPUT_DIRECTORY>

To process in the context of legacy data, the github option can be used with with --db flag, e.g.,

$ process_repo <DIR> --db github

NB If you have permission to write to the repository specified by the --repo flag, your changes will be committed back to the repo.

initialize_repo

This script migrates legacy SVN data in bulk. The root argument should point to the directory containing the contents of the SVN repository. The repodir should point to a directory where tool specific repositories should be created.

create_repo

This script is used to extract a given tool from the legacy SVN directory structure (whose directory is indicated by the root flag). The files for that tool will then be relocated to a vendor specific repository located at repo.

N.B. This script is not very useful. The first is because it is only used for migration and second because it migrates tools one at a time (vs. the initialize_repos scripts which will do them all at once). As such, it exists basically for historical purposes only.

Workflow

The overall workflow for processing FMI data from vendor supplied information all the way to support metric presented on the FMI web site is detailed in a separate document.