A simple word counter for more complex text processor files.

On command (be it manual or croned):

1. Retrieve list of known files from a database - waiting for database support
2. Crawl the directory structure from a given starting point - in progress
3. Count words in all supported files found in the directory - in progress
4. Update (or create) records in a database, keep history - waiting for database support
5. Display list of changed files and sum of the words - in progress

• Directory name - todo
• Specific files

• Changed database records
• Summary of changes on console

• The content is irrelevant; if a file lost 50 words but gained 150, the final result is to be increase by 100

• txt
• md
• docx
• odt

1. Linking walking the directories with processing
2. Better type recognition - docx vs. doc, accepting dots, etc.
3. Database handling
   1. Creating data structures
   2. Data retrieval
   3. Data saving
4. Incorrect filename handling
5. Counting algorithms verification - current results disagree with both Libre Office as with Microsoft Office
   • Note that LO disagrees with MO as well
6. Unit testing
   1. Mechanism
   2. Coverage for processor classes
7. Exchangeable database interface
   • May start with SQLite