Expert Needed: Clarion 5 / TopSpeed Data Extraction & Relational Key Mapping

What I have I have the data directory of an old Greek medical application built in Clarion (Clarion 5 — the install includes C5runx.dll, C5TPS.DLL, C5clax.dll and related library DLLs). It's a straightforward export job, not a big migration. The files - *.mfi — the data tables. Each starts with signature bytes C3, is XOR-encrypted, and holds Greek text in the windows-1253 codepage. - *.K01 … .K0E — the matching Clarion index/key files. - *.MEM — memo/long-text fields (signature M3, also XOR-encrypted). - I also have the original application EXE and library DLLs, which contain the table/field definitions. What I actually need I only need 4 specific tables exported — the patient table plus a few related clinical data tables. I do NOT need the whole database (it has ~200 tables); just these four, complete and correct. What's wrong right now I already have a working extraction attempt (a Python script) that decrypts the files and reads the data — so reading the format is basically solved. The problem is it doesn't reconstruct the relationships between tables: it can't reliably link a patient record to that patient's rows in the related data tables (the foreign-key / linking field isn't being matched correctly). So I get the data, but it isn't joined back to the right patient. So the task is: 1. Read these 4 tables correctly (fields, types, Greek text, memos). 2. Identify the key field(s) that link the patient table to the related tables, and join them properly so every data row is tied to the correct patient. 3. Export the result to clean CSV (or Excel/SQL — your call), correctly aligned and complete. Ideal experience - Hands-on with Clarion (Clarion 5 / TopSpeed) and its file drivers (.DAT / .TPS / .Kxx index files). Bonus for the TopSpeed ODBC driver or writing Clarion export utilities. - Comfortable identifying Clarion keys/relations and Greek (windows-1253) encoding. Important: This is medical data, so an NDA is required before any files are shared. I can provide redacted/sample files first so you can confirm you can read the format and the relations before we commit. Timeline: No hard deadline, but since this is a simple export I'd like it done fast. Deliverables: the 4 tables exported to CSV (or SQL), correctly joined patient-to-data, plus a short note on the method so it's reproducible. Please tell me your approach (Clarion itself / TopSpeed driver / custom parser) and a rough cost. This project is instantly paid upon completion.