Add SPDX analysis utilities#38
Conversation
|
I noted a few typo bugs in the code comments and the Usage text as I was preparing this pull request. These can be fixed with a follow-up patch if this is accepted. Note that it would be cleaner if 'esstra show' supported a '-f' option to show full paths instead of separate directories and files. That would eliminate the need for the 'esstra-full-paths' tool. |
Add some helper utilities to get information on the status of SPDX-License-Identifier lines in source files reported in esstra data. 'esstra-full-paths' is a brief utility to convert Directories and filenames in 'esstra show' output into full paths (used by has-spdx-id.py) This utility could possibly be replaced with a command line argument to the 'esstra' utility to show full paths. I started looking at that but the output handling in the 'esstra' utility was a bit more complex than I was expecting. This still might be the better way to handle this issue. 'has-spdx-id.py' is a utility to scan a file or list of files and report if they have SPDX-License-Identifier lines. It can also generate a count of files with and without the desired SPDX line. 'esstra-to-spdx-list.sh' is a program to take esstra data (as reported by 'esstra show', or as saved into a standalone file), and pipe it through standard Linux utilities and has-spdx-id.py to generate a report on the status of source files (whether they have SPDX-License-Identifier lines or not). These tools were used on the Linux kernel, using an Ubuntu (24.4) x86_64 kernel configuration for my Dell Desktop machine. The preliminary results for just the Linux kernel binary file (vmlinux at the top level directory) were: Files with SPDX lines: 6252 Files without SPDX lines: 577 Note that this omitted other files that would normally be associated with a full kernel built, such as all kernel module sources, and the entry and decompressor code used in a compressed kernel image (such as bzImage) Signed-off-by: Tim Bird <tim.bird@sony.com>
tbird20d
force-pushed
the
main
branch
2 times, most recently
from
07e88ca to
cb0274d
Compare
|
@tbird20d We agree with you that adding an option to As for the tools that check for |
3 participants
Add some helper utilities to get information on the status of SPDX-License-Identifier lines in source files reported in esstra data.
'esstra-full-paths' is a brief utility to convert Directories and filenames in 'esstra show' output into full paths (used by has-spdx-id.py) This utility could possibly be replaced with a command line argument to the 'esstra' utility to show full paths. I started looking at that but the output handling in the 'esstra' utility was a bit more complex than I was expecting. That still might be the better way to handle this issue.
'has-spdx-id.py' is a utility to scan a file or list of files and report if they have SPDX-License-Identifier lines. It can also generate a count of files with and without the desired SPDX line.
'esstra-to-spdx-list.sh' is a program to take esstra data (as reported by 'esstra show', or as saved into a standalone file), and pipe it through standard Linux utilities and has-spdx-id.py to generate a report on the status of source files (whether they have SPDX-License-Identifier lines or not).
These tools were used on the Linux kernel, using an Ubuntu (24.4) x86_64 kernel configuration for my Dell Desktop machine. The preliminary results for just the Linux kernel binary file (vmlinux at the top level directory) were:
Files with SPDX lines: 6252
Files without SPDX lines: 577
Note that this omitted other files that would normally be associated with a full kernel built, such as all kernel module sources, and the entry and decompressor code used in a compressed kernel image (such as bzImage)