Re: Cursor (second try)

Subject: Re: Cursor (second try)
From: sam th (
Date: Fri May 19 2000 - 14:08:09 CDT

Hash: SHA1

On Fri, 19 May 2000, Mike Nordell wrote:

> Please. Could someone change subject regarding "reply-to:" from Cursor to
> whatever you think is appropriate?
> Let's try this again.
> I *need* to know if the double linked list that fp_Run participate in
> should *always* only contain runs within one fp_Line. If no onw can
> answer this, I guess I just ditch the new cursor class.

Please don't ditch the cursor class. It sounds very promising. I've
tried to do some digging with LXR, but if you know this already, just
ignore it.

The linked list of fp_Run's is (in some nebulous sense) part of
fl_BlockLayout, in that fl_BlockLayout::m_pFirstRun indicates the first
item in the linked list. (For an example of this being used to traverse
the whole list, see fl_BlockLayout::__dump() ) fl_BlockLayout also
contains multiple fp_Line's (which form an additional linked list).
Therefore, it seems to me that there *need* be no correlation between the
fp_Run list and any particular fp_Line. Additionally, the __dump()
function descends the hierarchy rather nicely, and it goes from
fl_BlockLayout directly to fp_Run, bypassing fp_Line.

All of these factors combine to make me think that your 'strange' file was
not a bug. However, I have spent very little time in the formatter, so I
can't be at all sure of this. The original authors seem to be Paul and
Jeff, so they would be the authorities here.

hope this helps
                                     sam th
Version: GnuPG v1.0.1 (GNU/Linux)
Comment: For info see


This archive was generated by hypermail 2b25 : Fri May 19 2000 - 14:08:17 CDT