RFE: clarity about INI format

[mtbcarroll]

ome.scifio.common.IniParser parses INI files but its comment handling seems to mess up lines with a # before the = as part of the "key" name and it also doesn't handle multi-line values as are found for, say, the Recording #1 Notes key for zeiss-lsm/sample files.mdb/sample files.mdb [XY-ch] as imported under OMERO 4.

If we need to handle keys with a '#' and values with multiple lines, some blank, and maybe even '=' in the values and whatnot, then we need to make sure that the original_metadata.txt file generator code writes these files out properly in the first place, in accordance with a documented INI format that IniParser properly reads, even the edge cases. (I assume some of this becomes moot in OMERO 5, but even that could probably be handling the metadata for "legacy" images better, those that were imported before the upgrade to OMERO 5, and in passing metadata to Python or something that may still need a string-based format.)

Also note that in OMERO 5 OriginalMetadataRequestI newly has its own simple INI parser to avoid the issue with '#' in keys.

[mtbcarroll]

#10304 describes some parsing problems.

[jamoore]

Pushing.