Documentation: dev-tools: Clarify requirements for result description
Subsystems (3)
PATCH 1/1 Documentation: dev-tools: Clarify requirements for result description
Currently the KTAP specification says that a test result line is
<result> <number> [<description>][ # [<directive>] [<diagnostic data>]]
and the description of a test can be "any sequence of words (can't include #)" which specifies that there may be more than one word but does not specify anything other than those words which might be used to separate the words which probably isn't what we want. Given that practically we have tests using a range of separators for words including combinations of spaces and combinations of other symbols like underscores or punctuation let's just clarify that the description can contain any character other than # (marking the start of the directive/diagnostic) or newline (marking the end of this test result).
--- Documentation/dev-tools/ktap.rst | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
Documentation/dev-tools/ktap.rst +2 -2
@@ -80,8 +80,8 @@ have the number 1 and the number then must increase by 1 for each additional subtest within the same test at the same nesting level. The description is a description of the test, generally the name of -the test, and can be any string of words (can't include #). The -description is optional, but recommended. +the test, and can be any string of characters other than # or a +newline. The description is optional, but recommended. The directive and any diagnostic data is optional. If either are present, they must follow a hash sign, "#".
base-commit: 9abf2313adc1ca1b6180c508c25f22f9395cc780
Signature · 1 line
2.30.2
Quoted 39 lines
-----Original Message-----
From: Mark Brown <[email protected]>Currently the KTAP specification says that a test result line is
<result> <number> [<description>][ # [<directive>] [<diagnostic data>]]
and the description of a test can be "any sequence of words (can't include #)" which specifies that there may be more than one word but does not specify anything other than those words which might be used to separate the words which probably isn't what we want. Given that practically we have tests using a range of separators for words including combinations of spaces and combinations of other symbols like underscores or punctuation let's just clarify that the description can contain any character other than # (marking the start of the directive/diagnostic) or newline (marking the end of this test result).
Signed-off-by Mark Brown <[email protected]>--- Documentation/dev-tools/ktap.rst | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)Documentation/dev-tools/ktap.rst +2
-2@@ -80,8 +80,8 @@ have the number 1 and the number then must increase by 1 for each additional subtest within the same test at the same nesting level. The description is a description of the test, generally the name of -the test, and can be any string of words (can't include #). The -description is optional, but recommended. +the test, and can be any string of characters other than # or a +newline. The description is optional, but recommended. The directive and any diagnostic data is optional. If either are present, they must follow a hash sign, "#".
Quoted 3 lines
base-commit:
9abf2313adc1ca1b6180c508c25f22f9395cc780--
2.30.2
This change is OK by me.
Just a comment while we're on the subject of weird chars in the test description, I'll describe what Fuego does with these. In order to create a unique testcase globally unique identifier (TGUID), that can be used in databases, json, filenames, etc. Fuego converts spaces and other non-alphanumeric characters in the testcase description to underscores.
In particular, it's important for Fuego to distinguish between testcase identifiers and test suite identifiers (and Fuego uses '.' and '-') as delimiters and separators for those. The underscore conversion makes collisions possible, but unlikely when a testcase identifier is qualified with a test suite identifier.
It would be nice if we could get some consistency between LTP and kselftest (and the rest of the tests throughout the industry) on the delimiters used (and allowed) for different test element identifiers. Maybe the kernel has enough weight in the OSS ecosystem to set a standard here that others can follow.
-- Tim
Quoted 15 lines · Mark Brown
Currently the KTAP specification says that a test result line is
<result> <number> [<description>][ # [<directive>] [<diagnostic data>]]
and the description of a test can be "any sequence of words (can't include #)" which specifies that there may be more than one word but does not specify anything other than those words which might be used to separate the words which probably isn't what we want. Given that practically we have tests using a range of separators for words including combinations of spaces and combinations of other symbols like underscores or punctuation let's just clarify that the description can contain any character other than # (marking the start of the directive/diagnostic) or newline (marking the end of this test result).
Quoted 1 line
Signed-off-by Mark Brown <[email protected]>
Signature · 1 line
Kees Cook
Quoted 16 lines · Mark Brown
Currently the KTAP specification says that a test result line is
<result> <number> [<description>][ # [<directive>] [<diagnostic data>]]
and the description of a test can be "any sequence of words (can't include #)" which specifies that there may be more than one word but does not specify anything other than those words which might be used to separate the words which probably isn't what we want. Given that practically we have tests using a range of separators for words including combinations of spaces and combinations of other symbols like underscores or punctuation let's just clarify that the description can contain any character other than # (marking the start of the directive/diagnostic) or newline (marking the end of this test result).
Quoted 2 lines
Signed-off-by Mark Brown <[email protected]>---
Thanks for clarifying this: I totally agree that "sequence of words" wasn't a great description, and this is better.
+Rae in case she has any thoughts.
Cheers,
-- David