This python program uses Amazon S3 via 'boto' to read articles from the eLife article store and parse them for DOI and title information using the NLM parser, printing the result out in a format suitable for the alm server db:article:load call. Use --help for info on parameters.

These shell scripts work as a pair to save and restore the ALM database to a tar file stored in an S3 bucket. Along with the SQL and Couch databases, several of the more important settings files are saved (but not restored) in case of need.

ALMbackup configuration is all via variables in the code except for the AWS credentials needed to access the S3 bucket, which are expected to be in the config.json file of the ALM server (in /var/www/alm/shared).

ALMrestore uses the same credentials from config.json, and accepts a single command-line argument which is the name of the object to restore from, in the form: s3://bucket-name/yymm/yymmdd_hhmmsst.tgz (which is the format used to save data by backup).