go-fast | fast.com api for go - pure - no headless browser or stuff | REST library

Found what was happening. When pytessaract read the text a \x0c command were added in a new line after the last read line. Solved the issue by removing the command like this:

Each peoplecode type (as seen in the PSPROJECTITEM.OBJECTTYPE field) has its own particular mapping mechanism to the OBJECTID fields. The best way to proceed is open some of the code of interest and save it in a project. Then find the PSPCMPROG and PSPROJECTITEM rows associated with it and reason through the mapping at the OBJECTIDx and OBJECTVALUEy level.

For example:

• message pcode (objecttype 39) uses => AND M.OBJECTID1 = 60 and M.OBJECTID2 = 12.

• COMPONENT_INTERFACE_PEOPLECODE type (42) => M.OBJECTID1 = 74 and M.OBJECTID2 = 12. I assume (without checking) that OBJECTVALUE1 = PSBCDEFN.BCNAME, where Record.PSBCDEFN stores Component Interfaces.

Generally speaking expect the 1st key for a record to map to OBJECTVALUE1, the 2nd to OBJECTVALUE2...

Note also that some of the tables storing plaintext, not binary, pcode, are not kept in synch automatically - there's an AE batch that does this, typically used during upgrades. So make sure you are not pulling out stale data. And you may have to stitch long programs together with PROGSEQ.

That string is not Base-64 encoded. It is simply hexadecimal. (I have no idea why it ends in ==, which makes it look a little like a Base-64 string.) You should be able to see by inspection that there are no lower case letters, or for that matter upper case letters after F as there would be in a typical Base-64 encoded string of compressed, i.e. random-appearing data.

Remove the equal signs at the end and use .decode("hex") in Python 2, or bytes.fromhex() in Python 3.