Re: [RFC] redoxfs.pk: Pickle for RedoxFS file system

[Jose E. Marchesi]

> On Fri, Dec 18, 2020 at 06:10:06PM +0100, Jose E. Marchesi wrote:
>> 
>> > /* Node pretty-printer
>> >  *
>> >  * Prints the contents of a `RedoxFS_Node` in a readable fashion.
>> >  * To increase readability, only non-empty extents are get printed before 
>> > any
>> >  * other fields. By default, extents are get printed from first one to last
>> >  * one by default, but this can be changed by passing 1 in `reverse` 
>> > argument.
>> >  */
>> > fun redoxfs_nprint = (RedoxFS_Node n, int reverse = 0) void:
>> 
>> If you turn this into a proper method pretty printer (_print) in the
>> type RedoxFS_Node, then you can omit the `reverse' argument: see comment
>> below.
>> 
>> >   {
>> >     print ("#<\n  extents = [\n");
>> >
>> >     var f = reverse ? n.extents'length : 0UL;  /* first */
>> >     var l = reverse ? 0UL : n.extents'length;  /* last  */
>> >
>> >     while (f != l) {
>> >       if (reverse)
>> >         --f;
>> >
>> >       var e = n.extents[f];
>> >
>> >       if (!e.empty)
>> >         printf ("    .[%u64d]={block=%v, length=%v}\n", f, e.block, 
>> > e.length);
>> 
>> Looks like the above line should go to the pretty printer (_print) of
>> RedoxFS_Extent, including the handling of !empty.
>> 
>> The poke user could then do:
>> 
>> (poke) var fs = RedoxFS_Node @ WHEREVER
>> (poke) fs
>> #<
>>   extents=[#<...>, #<...>, ...],
>>   FILE
>>   ...
>> >
>
>
> The problem is the index part of the output: `.[%u64d]=`
> If we omit the empty blocks, the user needs a way to know which extents
> are get printed in the output.

But the index is not part of the extent itself... What contains extents?
Nodes, right?  Then I would say it is the node's pretty-printer who is
responsible to print these indexes...