Am 10.12.2011 01:49, schrieb Andreas Färber:
Am 10.12.2011 00:19, schrieb Stefan Weil:
This file can be used with codespell to suppress wrong
misspelling reports.
Signed-off-by: Stefan Weil <s...@weilnetz.de>
---
scripts/codespell.exclude | 215
+++++++++++++++++++++++++++++++++++++++++++++
1 files changed, 215 insertions(+), 0 deletions(-)
create mode 100644 scripts/codespell.exclude
diff --git a/scripts/codespell.exclude b/scripts/codespell.exclude
new file mode 100644
index 0000000..d69c946
--- /dev/null
+++ b/scripts/codespell.exclude
@@ -0,0 +1,215 @@
+ .rela.bss : { *(.rela.bss) }
+ .rela.bss : { *(.rela.bss) }
[...]
I don't like this.
Instead I suggest to keep a word list (that isn't whitespace-sensitive)
and to generate the .exclude file via shell script (grep -r) on the fly.
Andreas
grep -w. The solution which you suggest would need a list of filenames
and words because those words are only accepted in the given files.
The resulting exception list would also suppress real misspellings
of the given words in the given files which is not what we want.
If this seems acceptable, the exception list can be created using
a small shell script.
I don't say that an exception list is a nice solution (I also don't like
it),
but that's the way codespell works with the Linux kernel code, too.
Yes, the list is whitespace sensitive, but even if whitespace of the
affected
lines is changed, that's not a big problem. You will get some false
misspelling reports and either ignore them or update the exception list.
We could also change the code to reduce the need for spelling exceptions.
Renaming the variable aSign (which is reported as misspelling of assign)
would reduce the exception list a lot.
Regards,
Stefan
