* moved scripts into bin directory. Create `scanner` as a link to `main.py`. Added `make_album_ini.sh`to create a default `album.ini` file in a directory. * add options `social`, `social_size` and `social_color` for tuning display of social icons. * modified options names to a more logical naming: * - default_album_date_reverse_sort -> default_album_reverse_sort * - default_media_date_reverse_sort -> default_media_reverse_sort * use Debian/Ubuntu system-wide JavaScript packages if available (you might need to run `sudo a2enconf javascript-common` on the server to enable the use of `/javascript` virtual directory). * added support for `uglifyjs` JavaScript minifier * new option `small_square_crops_background_color` for filling the background of small square crops * bug fixes * new debug option `show_faces`: lets the scanner show the faces detected (but read the note in `myphotoshare.conf.default`) * new option `face_cascade_scale_factor`: a parameter of `opencv`'s `detectMultiScale` function that detects the faces in the photo * UI cleaned putting sort and view switches to a hidden top right menu * user can toggle album slide mode, thumbnails spacing, thumbnails types, show album and media names, show media count, all with right top corner menu * the scanner manages more precisely certain option changes that require regenerating of `json` files, reduced size images, thumbnails * new option `slide_album_caption_color`: the color to use with album slide mode * implemented search function: media and albums can be searched by file name, title, description, tags; search may be whole word or inside words, considering accents and capitals or not; search works with non-latin languages (e.g. oriental languages) too, but whole word searches will possibly give no result; stop words can be used in order reduce words bloat * new option `by_search_string`: the string used for search albums * new option `search_options_separator`: the character used for separating search options from search string in URIs * new option `max_search_album_number`: the maximum number of search album that will be loaded * new option `use_stop_words`: whether to use stop words when generating the search albums * removed `server_album_path` option and hard code it to `albums` * removed options `respected_processors` and substituded by `num_processors` option, defaults to 1: better for the user, does not oblige to set the value * reorganization of documentation: * creation of `doc` and `doc/img` folders * `README.md` split into individual files * explained how to use advanced features like geonames or face detection * created a gallery of screenshots * reduced images and thumbnail naming schema is now more robust * cache files names are now made of only lower case ascii characters * now photo metadata are read with three tools: * `Pillow's \_getexif()`, heritage of photofloat: good, but doesn't read the metadata of some old image * `exifread`: good tool, but crashes with some image which has exif data corrupted * `pyexiftool`, an `exiftool` wrapper: the most affordable, but the slowest The preference order for the tools is managed by the new option `metadata_tools_preference`