v1.5

• updated OpenRefine version (dev snapshot 2017-10-26 + pull request 1290)

v1.4

• replaced python client with one-file-executable (generated with pyinstaller for linux 64-bit)
• updated OpenRefine version (dev snapshot 2017-08-02)

v1.3

• use openrefine fork with added options for initial java heap space and autosave period
• set autosave period to 180 minutes

v1.2

• small performance improvement (ca. 3-5%) by changing the default initial heap space to max heap space.

v1.1

• now defaults to OpenRefine 2.7 stable (released 2017-06-18)
• added check if wget supports option --show-progress (since wget 1.16)
• added memoryload statistics to openrefine-batch-docker.sh

v1.0

Rebuilt the whole script to do everything with bash builtins. A variation of the new script that still uses docker is available as new file openrefine-batch-docker.sh.

This release uses OpenRefine v2.7 Release Candidate 2 (2017-03-03) by default.

additional improvements:

• getopts for parsing of arguments
• runtime tracking
• improved logging with additional if statement (skip transform and export steps if no config files are present and export is not set)
• improved cleanup: delete duplicates from copied projects

v0.6.1

• added support for empty directories

v0.6

• extended restart options for memory management: restartfile (restart docker after each project (e.g. input file)) and restarttransform (restart docker container after each transformation)
• added option to make export optional
• fixed bug in cross function (added a restart step to advertise copied projects to openrefine server)
• improved logging and documentation

v0.5

• fixed duplicate printing of openrefine server messages (caused by multiple calls of docker attach)
• silenced find error messages
• ps statistics now sorted
• always export files (even if no transformation rules are found)
• cross function input: replaced cp with rsync to exclude useless project-history

v0.4

• added option to restart server after each transformation