New

New compared to tv_grab_nl_py 2.2.x

Most changes are under the hood, but also some things to the user have changed.

• It should on average be faster as more heavy use is made of the cache. Only unknown days and the current day are fetched online. A first fetch on an empty cache might be a little slower. If for whatever reason you need an empty cache, don't delete the database-file. If however you do, first run with --configure as also information about the sources needed by some of them on every run is stored in there. You can use:

  • --clear-cache to remove all stored programming information. Outdated data is removed automatically.
  • --clear-ttvdb to remove all stored data from theTVdb.com.

  If you already have all data in the cache and do not want the program to refresh on today, you can use: --use-only-cache or -U.

• The merging of the sources is completely rewritten making it potentially more intelligent. So the output might differ a little as sometimes a different value gets preference. Previously it mostly took the first found value with in some cases like the credits adding new values from the other sources. Now choices are made after all data is in, also looking on disagreement between the sources at what value is given by most. Still the prime_source decides on equally divided votes. In time we might slightly update on this process.

• You now can use different languages for output/log messages. At present English and Dutch. I hope the number will grow as people supply them. Use language in your configuration or at the commandline to set it.

• If you are at a different then the default location you can change the timezone used for the output. Use output_tz in your configuration to set it.

• As all code now is generic, chances of recurring or hidden problems is slimmer. The differences between the sources is completely driven by the data from the source and the source files. While the code still has default actions, much more intensive is watched for exceptions to that default. Chances are that changes are automatically or through slight adaptation of the source file caught through similarity with other sources, leaving out the need for code adaptation (with added chance of regression errors) by us and upgrading by the user. This also will allow for more intensive testing. You will through your log receive messages when running with --configure is needed to update your configuration on changes in the source. Updates on the source-files are transparent. Also through your log you will receive messages on new releases. But once we have everything stable those won't be as often and potentially only to add new frontends and/or language-files. Those latter will probably be maxed to twice a year.

Added/changed Commandline/Config options

language

Use: Commandline, Configuration Section
Commandline: --language, -l
Values: Two letter language code. Currently "en" and "nl"
Default: "en"
As of version 1.0.0

Sets the language used (if available) for messages, log, etc. to the the given one. If a text is missing for that language, fall-back to the English text is tried, else an error is given. See in the .\tvgrabpyAPI\texts directory for the available language files. See languagefile on creating extra language files.

output_tz

Use: Configuration Section
Values: A string like "Europe/Amsterdam" or "UTC"
Default: "UTC" (depending on what your frontend has set)
As of version 1.0.0

With this option you determine what timezone is used in the xmltv output for start and stop times. See in Linux under /usr/share/zoneinfo for possible values. That path might differ slightly on your Linux flavor. Your frontend should set it to a reasonable default.
On the commandline you can always switch to "UTC" with the --utc or -u option. Originally in the configuration a similar boolean use_utc option is supported. It will be recognized, but use output_tz = UTC instead.

use-only-cache

Use: Commandline
Commandline: --use-only-cache, -U
Values: Boolean
As of version 1.0.0

Tels the API to not extract data from internet. It will only use already gathered data from the cache. This will make a run very fast.

ratingstyle

Use: Configuration Section
Values: long|short|single|none
Default: short
As of version 1.0.0

This was called "kijkwijzerstijl" in tv_grab_nl_py 2.2.x, a name that also will still be accepted. It defines how rating items are shown in the xmltv output.

• long : add the long descriptions and the icons
• short : add the one word descriptions and the icons
• single: add a single string (mythtv only reads the first item)
• none : don't add any