Algorithms » History » Version 19
Greg Burri, 08/07/2009 01:38 PM
| 1 | 1 | Greg Burri | h1. Algorithms |
|---|---|---|---|
| 2 | |||
| 3 | h2. Searching |
||
| 4 | 2 | Greg Burri | |
| 5 | 3 | Greg Burri | For a functional description see here : [[Functional definition#The-search-window]] |
| 6 | 15 | Greg Burri | |
| 7 | 1 | Greg Burri | !aybabtu_search.png! |
| 8 | 18 | Greg Burri | _This schema depicts how the results are sorted from one peer. Each peer result are then merged._ |
| 9 | 4 | Greg Burri | |
| 10 | 5 | Greg Burri | h2. Peer ID |
| 11 | |||
| 12 | 7 | Greg Burri | Each peer owns a peer id which is unique and generated during the first start. This ID is used to identify a peer, it's better than the previous usage of peer IP, considering this situation : |
| 13 | 5 | Greg Burri | |
| 14 | * _A_ put in queue a file entry _f_ from _B_, _B_ doesn't know the hashes of this file entry. |
||
| 15 | * _B_ change his IP address. |
||
| 16 | * _A_ want to download _f_, it can ask _B_ for the hashes even _B_'s IP changed. |
||
| 17 | |||
| 18 | 8 | Greg Burri | h2. Core threads |
| 19 | |||
| 20 | 9 | Greg Burri | There are three kind of threads in the core in addition to the main thread : |
| 21 | * [[Protocol_core-core#Downloading-thread|Downloading thread]] : @DownloadManager::ChunkDownloader@ |
||
| 22 | 8 | Greg Burri | * Uploading thread : @UploadManager::Uploader@ |
| 23 | 11 | Greg Burri | * [[Algorithms#Updating-the-file-cache|Updating file cache thread]] : @FileManager::FileUpdater@ |
| 24 | 5 | Greg Burri | |
| 25 | 19 | Greg Burri | h2. Updating the file cache (Obsolete, must be rewritten after the seventh prototype ended) |
| 26 | 10 | Greg Burri | |
| 27 | Here is the algorithm for the thread (@FileManager::FileUpdater@) which will periodically update the file cache and persist it. |
||
| 28 | |||
| 29 | <pre> |
||
| 30 | D : The set of shared directories |
||
| 31 | T : Time during the hashes are computed (for example 30s) |
||
| 32 | F : A set of file initialy empty |
||
| 33 | |||
| 34 | 13 | Greg Burri | - Add a watcher to the shared directories |
| 35 | |||
| 36 | // First synchronize (at start) |
||
| 37 | For each d in D (recursively) : |
||
| 38 | 14 | Greg Burri | - Synchronize physical folders and files with d content |
| 39 | 13 | Greg Burri | - Add in F the files which doesn't have computed hashes |
| 40 | |||
| 41 | Loop : |
||
| 42 | 10 | Greg Burri | t : Time.now |
| 43 | 1 | Greg Burri | For each f in F : |
| 44 | 13 | Greg Burri | - Compute the unknown hash of files f |
| 45 | 1 | Greg Burri | - Remove f from F |
| 46 | 10 | Greg Burri | If (Time.now - t) > T : break |
| 47 | - Wait T - (Time.now - t) |
||
| 48 | 13 | Greg Burri | - Wait for changes for a period of P |
| 49 | - When a modification occurs synchronise the file/folder |
||
| 50 | - Add each new file in F |
||
| 51 | 12 | Greg Burri | - Persist the entire cache in a file (only every ~30min) |
| 52 | 10 | Greg Burri | </pre> |
| 53 | |||
| 54 | 5 | Greg Burri | |
| 55 | 4 | Greg Burri | h2. Downloading |
| 56 | |||
| 57 | See here : [[Protocol_core-core#Downloading-threads]] |