You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The default is for now to run license, copyright and package. We need the info to always get a proper inventory of the codebase.
I am also thinking at eventually removing the CLI options to select only certain scans. The rationale is that several scans overlap:
For instance detecting an asserted license in a package manifest is best done by a combo of package detection + minimal license detection on the "license" field for that package. Rerunning license detection on the manifest as if it were a random, unstructured file is not productive nor accurate.
Same for copyrights part of the license text itself (e.g. in a GPL or LGPL for an FSF copyright) or present in a "copyright" field of a package manifest.
The text was updated successfully, but these errors were encountered:
I think at the very least, we should always default to -i.
I almost always want to use this flag during analysis work, and it is required by a number of our tools (AboutCode manager) and scripts.
I think it would be best to have -i-c and -l as defaults when running scancode without specific flags.
When specific flags are used (scancode -l file.txt output.json for example), scancode should do only what the user tells it. This is handy when running scancode on a specific license text or notice file, for example.
Sooner than later we need a way to store a set of options and we could provide an option file called "defaults". It is hard to anticipate all the different ways that people may want to use ScanCode so best to not choose for them, but make is easy for users to set their preferred options.
We can probably merge #209 with this issue.
The default is for now to run license, copyright and package. We need the info to always get a proper inventory of the codebase.
I am also thinking at eventually removing the CLI options to select only certain scans. The rationale is that several scans overlap:
For instance detecting an asserted license in a package manifest is best done by a combo of package detection + minimal license detection on the "license" field for that package. Rerunning license detection on the manifest as if it were a random, unstructured file is not productive nor accurate.
Same for copyrights part of the license text itself (e.g. in a GPL or LGPL for an FSF copyright) or present in a "copyright" field of a package manifest.
The text was updated successfully, but these errors were encountered: