XMLTV::Augment − Augment XMLTV listings files with automatic and user−defined rules.
Augment an XMLTV xml file by applying corrections ("fixups") to programmes matching defined criteria ("rules").
Two types of rules are actioned: (i) automatic, (ii) user-defined.
Automatic rules use pre-programmed input and output to modify the input programmes. E.g. removing a "title" where it is repeated in a "sub-title" (e.g. "Horizon" / "Horizon: Star Wars"), or trying to identify and extract series/episode numbers from the programme"s title, sub-title or description.
User-defined rules use the content of a "rules" file which allows programmes matching certain user-defined criteria to be corrected/enhanced with the user data supplied (e.g. adding/changing categories for all episodes of "Horizon", or fixing misspellings in titles, etc.)
By setting appropriate options in the "config" file, the "rules" file can be automatically downloaded using XMLTV::Supplement.
setEncoding
Set the assumed encoding of the rules file.
inputChannel
Store each channel found in the input programmes file for later processing by "stats".
augmentProgramme
Augment a programme using (i) pre-determined rules and (ii) user-defined rules. Which rules are processed is determined by the options set in the "config" file.
printInfo
Print the lists of actions taken and suggestions for further fixups.
end |
Do any final processing before exit (e.g. close the log file if necessary). |
new XMLTV::Augment( { ...parameters...} );
Possible
parameters:
rule => filename of file containing fixup rules (if
omitted then no user-defined rules will be actioned)
(overrides auto-fetch Supplement if that is defined; see
sample options file)
config => filename of config file to read (if omitted
then no config file will be used)
encoding => assumed encoding of the rules file (default =
UTF−8)
stats => whether to print the audit stats in the log
(values = 0,1) (default = 1)
log => filename of output log
debug => debug level (values 0−10) (default = no
debug)
note debug level > 3 is not likely to be of much use
unless you are developing code
1) Create the
XMLTV::Augment object
2) Pass each channel to inputChannel()
3) Pass each programme to augmentProgramme()
4) Tidy up using printInfo() & end()
#instantiate the object
my $augment = new XMLTV::Augment(
"rule" => "myrules.txt",
"config" => "myconfig.txt",
"log" => "augment.log",
);
die "failed to create XMLTV::Augment object" if
!$augment;
for each channel... {
# store the channel details
$augment−>inputChannel( $ch );
}
for each programme... {
# augmentProgramme will now do any requested processing of
the input xml
$prog = $augment−>augmentProgramme( $prog );
}
# log the stats
$augment−>printInfo();
# close the log file if necessary
$augment−>end();
Note: you are responsible for reading/writing to the XMLTV .xml file; the package will not do that for you.
remove_duplicated_new_title_in_ep
Rule #A1
Remove "New $title :" from <sub−title>
If
sub−title starts with "New" + <title>
+ separator, then it will be removed from the
sub−title
"separator" can be any of .,:;−
in : "Antiques Roadshow / New Antiques Roadshow:
Doncaster"
out: "Antiques Roadshow / Doncaster"
remove_duplicated_title_and_ep_in_ep
Rule #A2
Remove duplicated programme title *and* episode from <sub−title>
If
sub−title starts with <title> + separator +
<episode> + separator + <episode>, then it will
be removed from the sub−title
"separator" can be any of .,:;−
in : "Antiques Roadshow / Antiques Roadshow: Doncaster:
Doncaster"
out: "Antiques Roadshow / Doncaster"
remove_duplicated_title_in_ep
Rule #A3
Remove duplicated programme title from <sub−title>
i) If
sub−title starts with <title> + separator, then
it will be removed from the sub−title
ii) If sub−title ends with separator + <title>,
then it will be removed from the sub−title
iii) If sub−title starts with <title>(...), then
the sub−title will be set to the text in brackets
iv) If sub−title equals <title>, then the
sub−title will be removed
"separator" can be any of .,:;−
in : "Antiques Roadshow / Antiques Roadshow:
Doncaster"
out: "Antiques Roadshow / Doncaster"
in : "Antiques Roadshow / Antiques Roadshow
(Doncaster)"
out: "Antiques Roadshow / Doncaster"
in : "Antiques Roadshow / Antiques Roadshow"
out: "Antiques Roadshow / "
update_premiere_repeat_flags_from_desc
Rule #A4
Set the <premiere> element and remove any <previously−shown> element if <desc> starts with "Premiere." or "New series". Remove the "Premiere." text. Set the <previously−shown> element and remove any <premiere> element if <desc> starts with "Another chance" or "Rerun" or "Repeat"
check_potential_numbering_in_text
Rule #A5
Check for potential series, episode and part numbering in the title, episode and description fields.
extract_numbering_from_title
Rule #A5.1
Extract series/episode numbering found in <title>.
extract_numbering_from_episode
Rule #A5.2
Extract series/episode numbering found in <sub−title>.
extract_numbering_from_desc
Rule #A5.3
Extract series/episode numbering found in <desc>.
make_episode_from_part_numbers
Rule #A5.4
If no <sub−title> then make one from "part" numbers.
in :
"Panorama / " desc = "Part 1/2..."
out: "Panorama / Part 1 of 2"
process_user_rules
Rule #user
Process programme against user-defined fixups
The individual rules each have their own option to run or not; consider this like an on/off switch for all of them. I.e. if this option is off then no user rules will be run (irrespective of any other option flags).
process_non_title_info
Rule #1
Remove specified non-title text from <title>.
If title starts
with text + separator, then it will be removed from the
title
"separator" can be any of :;−
rule: 1|Python Night
in : "Python Night: Monty Python − Live at the
Hollywood Bowl / "
out: "Monty Python − Live at the Hollywood Bowl /
"
process_demoted_titles
Rule #11
Promote demoted title from <sub−title> to <title>.
If title
matches, and sub−title starts with text then remove
matching text from sub−title and move it into the
title.
Any text after 'separator' in the sub−title is
preserved. 'separator' can be any of .,:;−
rule: 11|Blackadder˜Blackadder II
in : "Blackadder / Blackadder II: Potato"
out: "Blackadder II / Potato"
process_replacement_titles_desc
Rule #10
Replace specified <title> / <sub−title> with title/episode pair supplied using <desc>.
If title &
sub−title match supplied data, then replace
<title> and <sub−title> with new data
supplied.
rule: 10|Which Doctor˜˜Gunsmoke˜Which
Doctor˜Festus and Doc go fishing, but are captured by a
family that is feuding with the Haggens.
in : "Which Doctor / " desc> = " Festus
and Doc go fishing, but are captured by a family that is
feuding with the Haggens. ..."
out: "Gunsmoke / Which Doctor"
process_replacement_titles_episodes
Rule #8
Replace specified <title> / <sub−title> with title/episode pair supplied.
If title &
sub−title match supplied data, then replace
<title> and <sub−title> with new data
supplied.
rule: 8|Top Gear USA Special˜Detroit˜Top
Gear˜USA Special
in : "Top Gear USA Special / Detroit"
out: "Top Gear / USA Special"
rule: 8|Top Gear USA Special˜˜Top Gear˜USA
Special
in : "Top Gear USA Special / "
out: "Top Gear / USA Special"
or
in : "Top Gear USA Special / 1/6."
out: "Top Gear / 1/6. USA Special"
process_mixed_title_subtitle
Rule #2
Extract sub-title from <title>.
If title starts
with text + separator, then the text after it will be moved
into the sub−title
"separator" can be any of :;−
rule: 2|Blackadder II
in : "Blackadder II: Potato / "
out: "Blackadder II / Potato"
process_mixed_subtitle_title
Rule #3
Extract sub-title from <title>.
If title ends
with separator + text, then the text before it will be moved
into the sub−title
"separator" can be any of :;−
rule: 3|Storyville
in : "Kings of Pastry :Storyville / "
out: "Storyville / Kings of Pastry"
process_reversed_title_subtitle
Rule #4
Reverse <title> and <sub−title>
If
sub−title matches the rule's text, then swap the title
and sub−title
rule: 4|Storyville
in : "Kings of Pastry / Storyville"
out: "Storyville / Kings of Pastry"
process_replacement_titles
Rule #5
Replace <title> with supplied text.
If title
matches the rule's text, then use the replacement text
supplied
rule: 5|A Time Team Special˜Time Team
in : "A Time Team Special / Doncaster"
out: "Time Team / Doncaster"
This is the one which you will probably use most. It can be used to fix most incorrect titles − e.g. spelling mistakes; punctuation; character case; etc.
process_subtitle_remove_text
Rule #13
Remove specified text from <sub−title> for a given <title>.
If
sub−title starts with text + separator, or ends with
separator + text,
then it will be removed from the sub−title.
"separator" can be any of .,:;− and is
optional.
rule: 13|Time Team˜A Time Team Special
in : "Time Team / Doncaster : A Time Team Special
"
out: "Time Team / Doncaster"
process_replacement_episodes
Rule #7
Replace <sub−title> with supplied text.
If
sub−title matches the rule's text, then use the
replacement text supplied
rule: 7|Time Team˜Time Team Special:
Doncaster˜Doncaster
in : "Time Team / Time Team Special: Doncaster"
out: "Time Team / Doncaster"
process_replacement_ep_from_desc
Rule #9
Replace <sub−title> with supplied text when the <desc> matches that given.
If
sub−title matches the rule's text, then use the
replacement text supplied
rule: 9|Heroes of Comedy˜The Goons˜The series
celebrating great British comics pays tribute to the Goons.
in : "Heroes of Comedy / "
out: "Heroes of Comedy / The Goons"
or
in : "Heroes of Comedy / Spike Milligan"
out: "Heroes of Comedy / The Goons"
process_replacement_genres
Rule #6
Replace <category> with supplied text.
If title
matches the rule's text, then use the replacement
category(−ies) supplied
(note ALL existing categories are replaced)
rule: 6|Antiques
Roadshow˜Entertainment˜Arts˜Shopping
in : "Antiques Roadshow / " category
"Reality"
out: "Antiques Roadshow / " category
"Entertainment" + "Arts" +
"Shopping"
You can specify a wildcard with the title by using %% which represents any number of characters. So for example "News%%" will match "News", "News and Weather", "Newsnight", etc. But be careful; "%%News%%" will also match "John Craven’s Newsround", "Eurosport News", "Election Newsroom Live", "Have I Got News For You", "Scuzz Meets Jason Newsted", etc.
process_replacement_film_genres
Rule #12
Replace "Film"/"Films" <category> with supplied text.
If title
matches the rule's text and the prog has category
"Film" or "Films", then use the
replacement category(−ies) supplied
(note ALL categories are replaced, not just
"Film")
rule: 12|The Hobbit
Special˜Entertainment˜Interview
in : "The Hobbit Special / " category
"Film" + "Drama"
out: "The Hobbit Special / " category
"Entertainment" + "Interview"
process_translate_genres
Rule #14
Replace <category> with supplied value(s).
If category
matches one found in the prog, then replace it with the
category(−ies) supplied
(note any other categories are left alone)
rule: 14|Soccer˜Football
in : "Leeds v Arsenal" category "Soccer"
out: "Leeds v Arsenal" category
"Football"
rule: 14|Adventure/War˜Action Adventure˜War
in : "Leeds v Arsenal" category
"Adventure/War"
out: "Leeds v Arsenal" category "Action
Adventure" + "War"
process_add_genres_to_channel
Rule #15
Add a category to all programmes on a specified channel.
If channel
matches this prog, the add the supplied category(−ies)
to the programme
(note any other categories are left alone)
rule: 15|travelchannel.co.uk˜Travel
in : "World's Greatest Motorcycle Rides" category
"Motoring"
out: "World's Greatest Motorcycle Rides" category
"Motoring" + "Travel"
rule: 15|cnbc.com˜News˜Business
in : "Investing in India" category ""
out: "Investing in India" category
"News" + "Business"
You should be very careful with this one as it will add the category you specify to EVERY programme broadcast on that channel. This may not be what you always want (e.g. Teleshopping isn’t really "music" even if it is on MTV!)
process_remove_numbering_from_programmes
Rule #16
Remove episode numbering from a given programme title (on an optionally-specified channel).
If title
matches the one in the prog, all programme numbering for the
programme
is removed, on any channel. An optional channel identifier
can be provided to
restrict the removal of programme numbering to the given
channel.
rule: 16|Bedtime Story
in : "CBeebies Bedtime Story" episode−num
".700."
out: "CBeebies Bedtime Story" episode−num
""
rule: 16|CBeebies Bedtime Story˜cbeebies.bbc.co.uk
in : "CBeebies Bedtime Story" episode−num
".700."
out: "CBeebies Bedtime Story" episode−num
""
Remember to specify the optional channel limiter if you have good programme numbering for a given programme title on some channels but not others.
Geoff Westcott, honir.at.gmail.dot.com, Dec. 2014.
This code is based on the "fixup" method/code defined in tv_grab_uk_rt grabber and credit is given to the author Nick Morrott.