Re: need help understanding underlying file format for cvs

[Mark D. Baushke]

Ronald Petty <ron.petty@unigeek.com> writes:

> I am starting developing some repository surfing tools, I know plenty of
> them are out there and most use rlog or something like that to do it. 
> In order to do what I want I need to know the underlying file format for
> the files.  I want to list history/user/date/the actual file based on
> the select version etc.
> I see the file,v has most of this, and I have read man rcsfile and it
> gives the grammer for an rcs file.  However I don't know if cvs uses
> this grammer for its file,v.  

cvs should be upward compatible with RCS format. The only thing that is
an extension is the use of a 'magic branch number' which uses a '.0.' as
the second rightmost position in a version number (ie, 1.1.0.2 is a
'magic branch number'). See the cvs documentation at

    http://www.cvshome.org/docs/manual/cvs_5.html#SEC59

for more details.

> So could someone tell me if this grammer is correct to develop a file
> parser/viewer.
> 
>        rcstext   ::=  admin {delta}* desc {deltatext}*

I believe that there must be one or more occurances of a {delta} element
in a cvs repository's representation of a ,v file. Likewise for the
deltatext being one or more occurances.
 
>        admin     ::=  head       {num};
>                       { branch   {num}; }
>                       access     {id}*;
>                       symbols    {sym : num}*;
>                       locks      {id : num}*;  {strict  ;}
>                       { comment  {string}; }
>                       { expand   {string}; }
>                       { newphrase }*

cvs does not use newphrase at the present time, but I think it may properly
ignore it...
 
>        delta     ::=  num
>                       date       num;
>                       author     id;
>                       state      {id};
>                       branches   {num}*;
>                       next       {num};
>                       { newphrase }*
> 
>        desc      ::=  desc       string
> 
>        deltatext ::=  num
>                       log        string
>                       { newphrase }*

Likewise. I think the newphrase may be supported enough to ignore it...

>                       text       string
> 
>        num       ::=  {digit | .}+
> 
>        digit     ::=  0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9
> 
>        id        ::=  {num} idchar {idchar | num}*
> 
>        sym       ::=  {digit}* idchar {idchar | digit}*

I believe that sym may NOT start with a digit.
 
>        idchar    ::=  any visible graphic character except special
> 
>        special   ::=  $ | , | . | : | ; | @
> 
>        string    ::=  @{any character, with @ doubled}*@
> 
>        newphrase ::=  id word* ;
> 
>        word      ::=  id | num | string | :
> 
> 
> Thanks for the help.
> Ron

Well, it is late here and I may not have looked closely enough, but I
think you may have overlooked RCS keywords in the text. However, that is
not really a grammer so much as how to intrepret the string provided by
the expand keyword under your <admin> token.

        Good luck,
        -- Mark