lynx-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: lynx-dev [PATCH] to dev13 - various fixes


From: Klaus Weide
Subject: Re: lynx-dev [PATCH] to dev13 - various fixes
Date: Sun, 31 Oct 1999 09:01:31 -0600 (CST)

On Sun, 31 Oct 1999, Vlad Harchev wrote:

> 
> * Bug in psrcmode code fixed. Now htmlsrc_tag encloses only tagname, not
> everything in generalized brackets <,>,</,/> - reported by KW

Note on detail: /> is not in general treated as a closing "generalized
bracket".  Not thus that it's not implemented (that too), it would
interfere with error recovery for some URLs that are improperly not
quoted: <A HREF=http://some.site/>.  Or to give an example where it
actually would make a difference, <BASE HREF=http://some.site/some/dir/>.

> * Bug with 'g' command fixed - reported by KW
> * default value for 'OPT' in SGML.c made '1', removed alternatives to code 
> surrounded by #if OPT1

Thank you very much.

>  Notes:
> * Seems that 'OPT' stuff doesn't contain errors. Let's enable it for several
> releases, and then remove alternative code.

Sounds good.

> * Added comments about pools.
> * Docs updated to reflect new functionality of -dont-wrap-pre


> @@ -3446,10 +3417,6 @@
>               start_element(context);
>  #ifdef USE_PSRC
>           if (psrc_view) {
> -             if (context->current_tag != context->unknown_tag)
> -                 PSRCSTOP(tag);
> -             else
> -                 PSRCSTOP(badtag);
>               PSRCSTART(abracket);
>               PUTC('>');
>               PSRCSTOP(abracket);

Nice to see some '-' lines. :)

> --- lynx2-8-3dev13u-was/lynx.cfg      Thu Oct 28 13:58:05 1999
> +++ lynx2-8-3dev13u/lynx.cfg  Sun Oct 31 17:13:36 1999
> @@ -2258,7 +2258,7 @@
>  #  Lexem                     LEXEMNAME FURTHER EXPLANATION

Btw. is it lexem or lexeme?

>  # ------------------------------------
>  # comment                    COMM
> -# tag                        TAG       everything in gen. brackets
> +# tag                        TAG       tag name only
                                          ^
Should that say "recognized tag"?

>  # attribute                  ATTRIB
>  # attribute value            ATTRVAL
>  # generalized brackets       ABRACKET  <  >  </

It doesn't list /> (and shouldn't).

> diff -ru lynx2-8-3dev13u-was/src/GridText.c lynx2-8-3dev13u/src/GridText.c
> --- lynx2-8-3dev13u-was/src/GridText.c        Fri Oct 29 18:26:35 1999
> +++ lynx2-8-3dev13u/src/GridText.c    Sun Oct 31 15:24:34 1999
> @@ -195,8 +195,61 @@
>      Can be 0 or 1*/
>  
>  /*these are generic macors for any pools (provided those structures have the 
> -    same members as HTStyleChangePool) - VH */
> -/*this macro sets 'ptr' to the "allocated" storage or NULL on failure.*/
> +same members as HTStyleChangePool). Pools are used for allocation of groups 
> +of objects of the same type T. Pools are represented as a list of structures
> +of type P (called pool chunks here). Structure P has an array of N objects 
> of 
> +type T named 'data' (the number N in the array can be chosen arbitrary), 
> +pointer to the next pool chunk named 'pool', and the number of free items 
> +in that pool chunk named 'free_items'. Ie, here is a definition of the
> +structure P:
> +     struct P
> +     {
> +         T data[N];
> +         struct P* next;
> +         int free_items;
> +     };
> + It's recommened that sizeof(P) be memory page size minus 32 in order 
> malloced
> +chunks to fit in machine page size.
> + Allocation of 'n' items in the pool is implemented as decrementing member 
> +'free_items' by 'n' if 'free_items' >= 'n', or allocating a new pool chunk 
> and
> +allocating 'n' items in that new chunk.

... or a combination?  (use up free_items, and allocate 'n - free_items' items
in new chunk)


   Klaus


reply via email to

[Prev in Thread] Current Thread [Next in Thread]