Untuk teman yg baca pesanku yg terakhir tentang zip-doc dan wikipedia offline, dan berminat main2 dengan kode: lihatlah git repositorinya... Kemarin malam aku refactor semua kode, sekarang jauh lebih jelas dan berguna. File formatnya juga sedikit ganti, tapi sekarang aku pikir dia sudah final.
dari irc konversasi dengan teman: and I'm really proud of my Wikipedia offline code... it started out as a mess of spaghetti code, and it has slowly improved and been factored... last night I finally rewrote the whole archive class, so now you have a free-standing archive class, with a Writer, a Reader and a compressor. It will accept arbitrary blobs with arbitrary names (and hard-links ), you can also set arbitrary metadata (which is the only part that is updateable afterwards), you flush, and it's all written to the index... You can specify if you want gzip or bzip2 encoding, and the size of the index - all this is autodetected by the Reader object, which can get .get blobs with arbitrary names (if they exist ) and get_meta ... 10:25 So it has nothing to do with Wikipedia or even websites... it could be used to store anything if you wanted it to... and all the actual dump script does is run through the files, use another freestanding function to compress and strip the html, add it as a blob - or as a hardlink if it's a redirect and flush... :) kodenya diambil dari repo - dari websitenya (http://repo.or.cz/w/zip-doc.git) juga ada opsi ambil "snapshot" versi terakhir, kalau tidak mau pake git langsung. aku akan upload file id.zdump dengan format yg baru sebentar lagi. stian [Non-text portions of this message have been removed]
