Boost logo

Boost :

From: Pavel Vozenilek (pavel_vozenilek_at_[hidden])
Date: 2004-09-12 14:47:30


"John Torjo" wrote:

> The FORMAL Review of "Output Formatter" library begins today,
>
I briefly looked on the library and wrote down few
notes. I am especially interested in point [4].

1. Name of the library would be better changed.
   Hearing it for the first time I suspected
   something like variant on std::endl.

2. The first part of documentation should
   be what is purpose of the library.

   I guess in this case it is debugging
   and testing support, 100% cases.

3. The name 'formatob' is not very good.
   "naryfmt" etc is horrible.

4. There seems to be significant overlap
   with Boost.Serialization (I suspect
   outfm doesn't handle cyclic
   structures).

   It may or may not be possible
   to implement outfm functionality
   with special archive.

   Was this considered?

5. To support debugging I would like to see
   many more features:

   - ability to wrap long lines in pretty way

   - ability to indent outputed
     data when they are part of complex structure

   - ability to generate HTML as output
     (+ ability to fold/unfold big data
      structures using Javascript).

   - helper function

       generate_html_output_and_show_it_in_browser()

   - ability to diff two outputed data
     where applicable and produce
     some easy to read report,
     possibly helper function:

generate_html_output_and_show_it_in_browser_together_with_diff_of_previous_d
ata_from_here()

   - some time ago John Torjo designed
     SMART_ASSERT library. If this will make
     into Boost, outfm would be perfect
     complement for it.

/Pavel


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk