Adapt scraper to openzim conventions

[benoit74]

The scraper needs to be adapted to match openzim conventions.

List below is maybe not yet exhaustive, this is what came into my mind while working on openzim/zimfarm#804.

• fix some issues around CLI parameters ; among others probably:

  • outpath CLI parameter is not taken into account
  • outzim is not relative to outpath and fallbacking to name, and does not support "{period}" placeholder
  • other dirs or not relative to outpath
  • more coherency among commands / phases of scraper operation

• add support for stats JSON file

• instead of multiple commands (fetch, prebuild, zim), have multiple phases in a single command, and flags to disable some phases if needed ; this will force to have more coherency among parameters of the various phases

• take into account new openZIM Python conventions (see https://github.com/openzim/_python-bootstrap, including wiki pages)

PS: @mdp : no worries, this is something we will do on our own (or help you with) ; this is usual on new scrapers and mostly linked to a WIP on our side to better explain our expectations (or not, and consider this issue is normal / to do on our side since very specific to our way of working)

[mdp]

Hey @benoit74, Yeah this all make sense, I can tackle most of this early next week. Thanks for the _python-bootstrap repo link, I'll try and line this repo up with it.

[mdp]

@benoit74 #12 is ready for review, although it doesn't address all the enhancements, it DRY's up the CLI arguments, and tries to match up to other python projects. It also should address #13

[benoit74]

I removed the "add support for stats JSON file" part because scraper is running so fast that it makes little sense to report progress.