[PATCH] Corrected syntax of ssl_conf_command and friends

[Sergey Kandaurov]

> On 21 Jun 2022, at 01:18, Maxim Dounin <mdounin at mdounin.ru> wrote:
> 
> Hello!
> 
> On Mon, Jun 20, 2022 at 11:33:43PM +0400, Sergey Kandaurov wrote:
> 
>> # HG changeset patch
>> # User Sergey Kandaurov <pluknet at nginx.com>
>> # Date 1655753604 -14400
>> # Mon Jun 20 23:33:24 2022 +0400
>> # Node ID 3e1af31b0ecd9e0ae5a8a93a15ff15cc0b1f98f6
>> # Parent 53734009b0c36276e66792e23180ee9665c95419
>> Corrected syntax of ssl_conf_command and friends.
>> 
>> Each directive takes exactly two arguments: command name and value.
>> The parameter names follow SSL_CONF_cmd(3) documentation of OpenSSL 3.0.
>> 
>> diff --git a/xml/en/docs/http/ngx_http_grpc_module.xml b/xml/en/docs/http/ngx_http_grpc_module.xml
>> --- a/xml/en/docs/http/ngx_http_grpc_module.xml
>> +++ b/xml/en/docs/http/ngx_http_grpc_module.xml
>> @@ -10,7 +10,7 @@
>> <module name="Module ngx_http_grpc_module"
>> link="/en/docs/http/ngx_http_grpc_module.html"
>> lang="en"
>> - rev="7">
>> + rev="8">
>> 
>> <section id="summary">
>> 
>> @@ -535,7 +535,7 @@ The full list can be viewed using the
>> 
>> 
>> <directive name="grpc_ssl_conf_command">
>> -<syntax><value>command</value> ...</syntax>
>> +<syntax><value>option</value> <value>value</value></syntax>
> 
> I don't think such wording improves readability, since it 
> basically contradicts the "command" in the name of the directive.

Yeah, that's true.

> 
> The OpenSSL 3.0 documentation of SSL_CONF_cmd() does not seem to 
> be self-consistent, and uses terms "command", "operation" and 
> "option" in various places. Further, the term "option" is already 
> used by SSL_CTX_set_options() in a completely different meaning, 
> and it is also configurable with SSL_CONF_cmd("Options").
> 
> Overall, I would rather suggest to stick with the
> 
> ssl_conf_command name value;
> 
> syntax, which seems to be much more consistent and 
> self-explanatory.
> 
> [...]

Committed with adjustments, thanks.

-- 
Sergey Kandaurov