Enhancements to the ARSDOC Commandline API

Hi, we think some enhancements to the ARSDOC CLI API would be helpful to anyone who uses it.  We've noticed the behavior is slightly differently than the Desktop Application and ODWEK API, it also has a few things that with slight tweaks could help make extracting content easier and more complete.

a) When using the -f argument to pass in a folder name while not specifying the -G (application group) argument, the where clause is still expected to use a DB field name rather than folder field name. This is problematic when there are multiple AG's mapped and their AG field names are inconsistent, for example if the folder field name is BranchNumber, and AG1 has a mapped field called Branch, AG2 has a mapped field called Branch, and AG3 has a mapped field called BranchNum.  In the where clause, it's expected to pass the AG field name, so the query will break unless we run three separate arsdoc get commands while using the -G argument and specifying the correct AG field name for each AG in the where clauses. This works differently than OnDemand application itself or CNAV/ODWEK where you can query the folder field when querying the folder.

b) You can either use AG field names or your own string to name the output files, but not a combination of both. For example, I ARSDOC GET won't allow us to name the output files using metadata and adding '.TXT' to the end of the output filename(s). ie: -o "(accountno)(branchno).TXT" won't work and neither will -o "(accountno)-(branchno)". This would be nice to be able to have OnDemand name a file based on metadata but also be able to add custom text to parts of the filename.

c) when using the option to have output files named from metadata values if there's a '/' in the value, most notably in date fields, the command can fail or sometimes Linux tries to build a directory structure where the '/' creates a subdirectory. This makes adding date values as part of the output file names difficult unless the dates are formatted without /'s.

d) When using fieldnames to create output filenames using metadata values, if there are multiple documents sharing the same metadata values, each subsequent document overwrites the previous.  It would be great if a sequence number or uniquifier got appended, similarly to when you don't tell OnDemand to name the files using metadata values.  For example, if we use -o "(branch)(account)" and there are 5 documents with the same branch and account number, I will only get a single document created. ARSDOC GET essentially retrieves all of the documents, one a a time, but each document retrieved overwrites the previous file rather than concatenating them together or adding a unique string to the ends of the filenames. I believe it would be better if OnDemand did something like, <branch>.<account>.1 <branch>.<account>.2 etc etc, similarly to how files are generated when using -o without metadata field names.

Thanks for your consideration!

Dan