Episode numbers missing
Jeremy Nicoll - ml get_iplayer
jn.ml.gti.91 at wingsandbeaks.org.uk
Sun Mar 8 05:38:43 PDT 2015
"Dave Liquorice" <allsorts at howhill.com> wrote:
>Of the series I routinely collect Horizon is the worst.
Well, Storyville is bad too.
In my own program which I use to keep track of what I've downloaded (or
skipped - the download history's no good for that), entries for various
programmes need changed after the v2.91 change in data layouts in the cache
files, but even before that my code for Horizon programmes typically
contains lines like:
call omit "\|2013-2014: Dinosaurs: 4. The Hunt for Life|\"
call omit "\|2013-2014: 8. The Power of the Placebo|\"
call omit "\|2013-2014: 9. How You Really Make Decisions|\"
call omit "\|2013-2014: Sugar v Fat|\"
call omit "\|2013-2014: 11. Where is Flight MH370?|\"
call omit "\|2013-2014: 12. What's Wrong with Our Weather?|\"
(these follow some code which mean at that point the program is only looking
at cache lines which contained either: '|tv|Horizon|' or '|tv|Horizon:'.)
Now that v2.91 makes "Horizon: 2013-2014" a separate list of programmes from
those for, for example, "Horizon: 2012-2013" I could change my program to
treat them as separate groups of episodes. But then, maybe, I'd need to
change that back if the way the BBC list things, and hence how GiP creates
cache files, changes again. I've not decided yet.
Increasingly I code 'pid' in my tests, because although episode names and
numbers change format, and descriptive texts sometimes get changed when a
program is retransmitted, pids are fixed. So I may have lines like:
call omit "\Series 2|\ & \|b04ttdk7|\ & \|6. Mortal Coil|\"
call omit "\Series 2|\ & \|b04tthb7|\ & \|7. The Silence|\"
ie exclude from the list of episodes that I may be interested in, any that
include all three of those text snippets. Previously I'd also delete all
the rules I had for separate episodes of something when a series ended, ie
replace all the lines like those above with:
call omit "\Series 2|\"
but now, for some programmes, I keep the multiple rules because the
information associating pid & episode name may help clarify things in
For programmes which don't have episode names, I do generally try to put
some snippet of descriptive text into my 'omit' rules, eg:
call prog "Helen Keen's It Is Rocket Science"
call omit "\Series 1|\ & \Episode 1|\ &
\rockets and the men, women\"
call omit "\Series 1|\ & \Episode 2|\ & \von Braun - from Nazi to\"
call omit "\Series 3|\ & \Episode 2|\ & \obsession with UFOs\"
call omit "\Series 3|\ & \Episode 3|\ & \greats of astronomy\"
call omit "\Series 3|\ & \Episode 4|\ & \Astronomical Errors\"
because it reduces the risk of a single typo of eg an episode number (by me
in my rules, or by the BBC when they list something) from meaning a rule
ignores a programme I was interested in.
The BBC are sometimes lazy though in how they list stuff; whereas University
Challenge used to include in its description which teams were playing, all
the recent episodes have just had generic values.
Having my own logic to tell me which programmes listed in a cache file might
be of interest means I can do things like:
if yyyymm == "201503" then call omit "\Series 4|\"
for some programme; that is, I've not downloaded them, and I don't want to
be told about them any time this month (too busy writing programs to keep
track of programmes to actually watch any!) but next month I may well want
to know about them...) and it also means I can keep comments about why I
liked or hated some programme in amongst the rules.
My program's quite complicated; I can find things in terms of snippets of
subject or presenters' names or anything else, which means I'll always know
eg if a presenter whose style I like starts a new programme I've never heard
of (provided their name is in the programme name, episode name or the
description). For example I search for "James May" but exclude from those
hits listings of "Top Gear". Separately I search for "Top Gear" episodes
because once in a blue moon I watch one. I end up with a list of items
which might be of interest because they match a whole set of different
tests, and might well have been excluded (in theory) for a bunch of other
reasons, but if any one reason is still left at the end of that, the thing
Jeremy Nicoll - my opinions are my own.
More information about the get_iplayer