Re: [PATCH] Add Documentation/CodingStyle

Johannes Schindelin [off-list ref] writes:

I am not sure if we want to have CodingStyle document, but the
above are not CodingStyle issues.

If we are to write this down, I'd like to have the more
important third rule, which is:

 - In spite of the above two rules, we sometimes say "Although
   this is not in POSIX, it (is so convenient | makes the code
   much more readable | has other good characteristics) and
   practically all the platforms we care about support it, so
   let's use it".  Again, we live in the real world, and it is
   sometimes a judgement call, decided based more on real world
   constraints people face than what the paper standard says.

+For C programs:
+
+ - Use tabs to increment, and interpret tabs as taking up to 8 spaces

What's the character for decrement?  DEL? ;-)

+   Double negation is often harder to understand than no negation at
+   all.
+
+   Some clever tricks, like using the !! operator with arithmetic
+   constructs, can be extremely confusing to others.  Avoid them,
+   unless there is a compelling reason to use them.

I actually think (!!var) idiom is already established in our
codebase.

+ - Use the API.  No, really.  We have a strbuf (variable length string),
+   several arrays with the ALLOC_GROW() macro, a path_list for sorted
+   string lists, a hash map (mapping struct objects) named
+   "struct decorate", amongst other things.

 - When you come up with an API, document it.

+ - if you are planning a new command, consider writing it in shell or
+   perl first, so that changes in semantics can be easily changed and
+   discussed.  Many git commands started out like that, and a few are
+   still scripts.

No Python allowed?

Junio C Hamano wrote:

Johannes Schindelin [off-list ref] writes:

quoted

diff --git a/Documentation/CodingStyle b/Documentation/CodingStyle
new file mode 100644
index 0000000..622b80b
--- /dev/null
+++ b/Documentation/CodingStyle

@@ -0,0 +1,87 @@
+As a popular project, we also have some guidelines to keep to the
+code.  For git in general, two rough rules are:
+
+ - Most importantly, we never say "It's in POSIX; we'll happily
+   screw your system that does not conform."  We live in the
+   real world.
+
+ - However, we often say "Let's stay away from that construct,
+   it's not even in POSIX".
+

I am not sure if we want to have CodingStyle document, but the
above are not CodingStyle issues.

If we are to write this down, I'd like to have the more
important third rule, which is:

 - In spite of the above two rules, we sometimes say "Although
   this is not in POSIX, it (is so convenient | makes the code
   much more readable | has other good characteristics) and
   practically all the platforms we care about support it, so
   let's use it".  Again, we live in the real world, and it is
   sometimes a judgement call, decided based more on real world
   constraints people face than what the paper standard says.

quoted

+For C programs:
+
+ - Use tabs to increment, and interpret tabs as taking up to 8 spaces

What's the character for decrement?  DEL? ;-)

quoted

+   Double negation is often harder to understand than no negation at
+   all.
+
+   Some clever tricks, like using the !! operator with arithmetic
+   constructs, can be extremely confusing to others.  Avoid them,
+   unless there is a compelling reason to use them.

I actually think (!!var) idiom is already established in our
codebase.

Not very widely for arithmetic expressions though. I believe this
alludes to the (!!a + !!b + !!c) idiom used earlier, where it's not
exactly clear *why* a, b and c can be > 1 if != 0 is the only thing
we care about.

quoted

+ - Use the API.  No, really.  We have a strbuf (variable length string),
+   several arrays with the ALLOC_GROW() macro, a path_list for sorted
+   string lists, a hash map (mapping struct objects) named
+   "struct decorate", amongst other things.

 - When you come up with an API, document it.

quoted

+ - if you are planning a new command, consider writing it in shell or
+   perl first, so that changes in semantics can be easily changed and
+   discussed.  Many git commands started out like that, and a few are
+   still scripts.

No Python allowed?

Perhaps with this addendum?

- Think very, very hard before introducing a new dependency into git. This
  means you should stay away from scripting languages not already used in
  the git core command set unless your command is clearly separate from it,
  such as an importer to convert random-scm-X repositories to git.

-- 
Andreas Ericsson                   andreas.ericsson@op5.se
OP5 AB                             www.op5.se
Tel: +46 8-230225                  Fax: +46 8-230231

Hi,

On Tue, 6 Nov 2007, Junio C Hamano wrote:

Johannes Schindelin [off-list ref] writes:

quoted

diff --git a/Documentation/CodingStyle b/Documentation/CodingStyle
new file mode 100644
index 0000000..622b80b
--- /dev/null
+++ b/Documentation/CodingStyle

@@ -0,0 +1,87 @@
+As a popular project, we also have some guidelines to keep to the
+code.  For git in general, two rough rules are:
+
+ - Most importantly, we never say "It's in POSIX; we'll happily
+   screw your system that does not conform."  We live in the
+   real world.
+
+ - However, we often say "Let's stay away from that construct,
+   it's not even in POSIX".
+

I am not sure if we want to have CodingStyle document, but the
above are not CodingStyle issues.

Would you like to call it CodingConventions?

If we are to write this down, I'd like to have the more
important third rule, which is:

 - In spite of the above two rules, we sometimes say "Although
   this is not in POSIX, it (is so convenient | makes the code
   much more readable | has other good characteristics) and
   practically all the platforms we care about support it, so
   let's use it".  Again, we live in the real world, and it is
   sometimes a judgement call, decided based more on real world
   constraints people face than what the paper standard says.

Okay, will add.

quoted

+For C programs:
+
+ - Use tabs to increment, and interpret tabs as taking up to 8 spaces

What's the character for decrement?  DEL? ;-)

Hehe.  As you undoubtedly guessed, I meant "indent"...

quoted

+   Double negation is often harder to understand than no negation at
+   all.
+
+   Some clever tricks, like using the !! operator with arithmetic
+   constructs, can be extremely confusing to others.  Avoid them,
+   unless there is a compelling reason to use them.

I actually think (!!var) idiom is already established in our
codebase.

Yep, but when there are easier variants, AFAICT they were preferred.

quoted

+ - Use the API.  No, really.  We have a strbuf (variable length string),
+   several arrays with the ALLOC_GROW() macro, a path_list for sorted
+   string lists, a hash map (mapping struct objects) named
+   "struct decorate", amongst other things.

 - When you come up with an API, document it.

Okay.

quoted

+ - if you are planning a new command, consider writing it in shell or
+   perl first, so that changes in semantics can be easily changed and
+   discussed.  Many git commands started out like that, and a few are
+   still scripts.

No Python allowed?

Well, maybe we should allow that again.  Although I hoped we are over that 
now, as it would complicate the msysGit efforts tremendously.

Ciao,
Dscho

Even if our code is quite a good documentation for our coding style,
some people seem to prefer a document describing it.

The part about the shell scripts is clearly just copied from one of
Junio's helpful mails, and some parts were added from comments by
Junio and Andreas Ericsson.

Signed-off-by: Johannes Schindelin <redacted>
---

	On Wed, 7 Nov 2007, Andreas Ericsson wrote:

	> Perhaps with this addendum?
	> 
	> - Think very, very hard before introducing a new dependency into git. This
	>  means you should stay away from scripting languages not already used in
	>  the git core command set unless your command is clearly separate from it,
	>  such as an importer to convert random-scm-X repositories to git.

	I edited it minimally.

 Documentation/CodingStyle |  103 +++++++++++++++++++++++++++++++++++++++++++++
 1 files changed, 103 insertions(+), 0 deletions(-)
 create mode 100644 Documentation/CodingStyle

diff --git a/Documentation/CodingStyle b/Documentation/CodingStyle
new file mode 100644
index 0000000..38a3d9f
--- /dev/null
+++ b/Documentation/CodingStyle

@@ -0,0 +1,103 @@
+As a popular project, we also have some guidelines to keep to the
+code.  For git in general, two rough rules are:
+
+ - Most importantly, we never say "It's in POSIX; we'll happily
+   screw your system that does not conform."  We live in the
+   real world.
+
+ - However, we often say "Let's stay away from that construct,
+   it's not even in POSIX".
+
+ - In spite of the above two rules, we sometimes say "Although
+   this is not in POSIX, it (is so convenient | makes the code
+   much more readable | has other good characteristics) and
+   practically all the platforms we care about support it, so
+   let's use it".  Again, we live in the real world, and it is
+   sometimes a judgement call, decided based more on real world
+   constraints people face than what the paper standard says.
+
+
+As for more concrete guidelines, just imitate the existing code
+(this is a good guideline, no matter which project you are contributing
+to...).  But if you must have some list of rules, here they are.
+
+For shell scripts specifically (not exhaustive):
+
+ - We prefer $( ... ) for command substitution; unlike ``, it
+   properly nests.  It should have been the way Bourne spelled
+   it from day one, but unfortunately isn't.
+
+ - We use ${parameter-word} and its [-=?+] siblings, and their
+   colon'ed "unset or null" form.
+
+ - We use ${parameter#word} and its [#%] siblings, and their
+   doubled "longest matching" form.
+
+ - We use Arithmetic Expansion $(( ... )).
+
+ - No "Substring Expansion" ${parameter:offset:length}.
+
+ - No shell arrays.
+
+ - No strlen ${#parameter}.
+
+ - No regexp ${parameter/pattern/string}.
+
+ - We do not use Process Substitution <(list) or >(list).
+
+ - We prefer "test" over "[ ... ]".
+
+ - We do not write noiseword "function" in front of shell
+   functions.
+
+For C programs:
+
+ - Use tabs to indent, and interpret tabs as taking up to 8 spaces
+
+ - Try to keep to 80 characters per line
+
+ - When declaring pointers, the star sides with the variable name, i.e.
+   "char *string", not "char* string" or "char * string".  This makes
+   it easier to understand "char *string, c;"
+
+ - Do not use curly brackets unnecessarily.  I.e.
+
+	if (bla) {
+		x = 1;
+	}
+
+   is frowned upon.  A gray area is when the statement extends over a
+   few lines, and/or you have a lengthy comment atop of it.
+
+ - Try to make your code understandable.  You may put comments in, but
+   comments invariably tend to stale out when the code they were
+   describing changes.  Often splitting a function into two makes the
+   intention of the code much clearer.
+
+   Double negation is often harder to understand than no negation at
+   all.
+
+   Some clever tricks, like using the !! operator with arithmetic
+   constructs, can be extremely confusing to others.  Avoid them,
+   unless there is a compelling reason to use them.
+
+ - Use the API.  No, really.  We have a strbuf (variable length string),
+   several arrays with the ALLOC_GROW() macro, a path_list for sorted
+   string lists, a hash map (mapping struct objects) named
+   "struct decorate", amongst other things.
+
+ - When you come up with an API, document it.
+
+ - #include system headers in git-compat-util.h.  Some headers on some
+   systems show subtle breakages when you change the order, so it is
+   best to keep them in one place.
+
+ - if you are planning a new command, consider writing it in shell or
+   perl first, so that changes in semantics can be easily changed and
+   discussed.  Many git commands started out like that, and a few are
+   still scripts.
+
+ - Avoid introducing a new dependency into git. This means you usually
+   should stay away from scripting languages not already used in the git
+   core command set (unless your command is clearly separate from it,
+   such as an importer to convert random-scm-X repositories to git).

-- 
1.5.3.5.1597.g7191

onsdag 07 november 2007 skrev Johannes Schindelin:

Can we use less offensive wording in documentation than we do on the list or IRC?
I'm not hurt by it but it doesn't look serious.

+ - Try to keep to 80 characters per line

less than?

+
+ - When declaring pointers, the star sides with the variable name, i.e.
+   "char *string", not "char* string" or "char * string".  This makes
+   it easier to understand "char *string, c;"

Rationale: The C syntax is defined the way it is. C programmers should
understand it.

I'd amend: Don't mix different types in declarations.

Write:
	char *string;
	char c;

Rather than:

	char *string,c;

+ - Do not use curly brackets unnecessarily.  I.e.
+
+	if (bla) {
+		x = 1;
+	}
+
+   is frowned upon.  A gray area is when the statement extends over a
+   few lines, and/or you have a lengthy comment atop of it.

Avoid unnecessary curly brackets but use them if unsure. (This is less 
strict than "do not". There are cases where we do want "unnecessary"
brackets.

-- robin

Even if our code is quite a good documentation for our coding style,
some people seem to prefer a document describing it.

The part about the shell scripts is clearly just copied from one of
Junio's helpful mails, and some parts were added from comments by
Junio, Andreas Ericsson and Robin Rosenberg.

Signed-off-by: Johannes Schindelin <redacted>
---

	I think I owe this list an apology for starting yet another
	thread which seems to instigate comments by not many
	code contributors.

	My intention was to make things simpler, but it appears
	that the human brain was created such that it needs complexity
	and creates it when it is absent.

	For example, when I wrote "guidelines" I fully expected that
	it was understood that these are no natural laws which you
	cannot break, should the circumstances afford it.

	Given the discussion, I am half convinced that even this
	patch is a bad idea, and am quite willing to just stop sending
	updates to it.

 Documentation/CodingGuidelines |  106 ++++++++++++++++++++++++++++++++++++++++
 1 files changed, 106 insertions(+), 0 deletions(-)
 create mode 100644 Documentation/CodingGuidelines

diff --git a/Documentation/CodingGuidelines b/Documentation/CodingGuidelines
new file mode 100644
index 0000000..2d8656f
--- /dev/null
+++ b/Documentation/CodingGuidelines

@@ -0,0 +1,106 @@
+As a popular project, we also have some guidelines to keep to the
+code.  For git in general, two rough rules are:
+
+ - Most importantly, we never say "It's in POSIX; we'll happily
+   ignore your needs should your system that does not conform."
+   We live in the real world.
+
+ - However, we often say "Let's stay away from that construct,
+   it's not even in POSIX".
+
+ - In spite of the above two rules, we sometimes say "Although
+   this is not in POSIX, it (is so convenient | makes the code
+   much more readable | has other good characteristics) and
+   practically all the platforms we care about support it, so
+   let's use it".  Again, we live in the real world, and it is
+   sometimes a judgement call, decided based more on real world
+   constraints people face than what the paper standard says.
+
+
+As for more concrete guidelines, just imitate the existing code
+(this is a good guideline, no matter which project you are contributing
+to...).  But if you must have some list of rules, here they are.
+
+For shell scripts specifically (not exhaustive):
+
+ - We prefer $( ... ) for command substitution; unlike ``, it
+   properly nests.  It should have been the way Bourne spelled
+   it from day one, but unfortunately isn't.
+
+ - We use ${parameter-word} and its [-=?+] siblings, and their
+   colon'ed "unset or null" form.
+
+ - We use ${parameter#word} and its [#%] siblings, and their
+   doubled "longest matching" form.
+
+ - We use Arithmetic Expansion $(( ... )).
+
+ - No "Substring Expansion" ${parameter:offset:length}.
+
+ - No shell arrays.
+
+ - No strlen ${#parameter}.
+
+ - No regexp ${parameter/pattern/string}.
+
+ - We do not use Process Substitution <(list) or >(list).
+
+ - We prefer "test" over "[ ... ]".
+
+ - We do not write noiseword "function" in front of shell
+   functions.
+
+For C programs:
+
+ - Use tabs to indent, and interpret tabs as taking up to 8 spaces
+
+ - Try to keep to at most 80 characters per line
+
+ - When declaring pointers, the star sides with the variable name, i.e.
+   "char *string", not "char* string" or "char * string".  This makes
+   it easier to understand "char *string, c;"
+
+ - Do not use curly brackets unnecessarily.  I.e.
+
+	if (bla) {
+		x = 1;
+	}
+
+   is frowned upon.  A gray area is when the statement extends over a
+   few lines, and/or you have a lengthy comment atop of it.  Also,
+   like in the Linux kernel, if there is a long list of "else if"
+   statements, it can make sense to add curly brackets to single
+   line blocks.
+
+ - Try to make your code understandable.  You may put comments in, but
+   comments invariably tend to stale out when the code they were
+   describing changes.  Often splitting a function into two makes the
+   intention of the code much clearer.
+
+   Double negation is often harder to understand than no negation at
+   all.
+
+   Some clever tricks, like using the !! operator with arithmetic
+   constructs, can be extremely confusing to others.  Avoid them,
+   unless there is a compelling reason to use them.
+
+ - Use the API.  No, really.  We have a strbuf (variable length string),
+   several arrays with the ALLOC_GROW() macro, a path_list for sorted
+   string lists, a hash map (mapping struct objects) named
+   "struct decorate", amongst other things.
+
+ - When you come up with an API, document it.
+
+ - #include system headers in git-compat-util.h.  Some headers on some
+   systems show subtle breakages when you change the order, so it is
+   best to keep them in one place.
+
+ - if you are planning a new command, consider writing it in shell or
+   perl first, so that changes in semantics can be easily changed and
+   discussed.  Many git commands started out like that, and a few are
+   still scripts.
+
+ - Avoid introducing a new dependency into git. This means you usually
+   should stay away from scripting languages not already used in the git
+   core command set (unless your command is clearly separate from it,
+   such as an importer to convert random-scm-X repositories to git).

-- 
1.5.3.5.1597.g7191