TOPIC: how to save, retain and retrieve files as they change. There seem to be several approaches and one proposal here involved using altered file names with a numeric suffix.
Can we compare two different concepts? A typographical approach with little or no built-in support like searching for a number at the end of a file name and incrementing it, is very different than older approaches. VMS had file names with three fixed-length parts in some kind of struct as stored in the file system. The main file name had a maximum length. The file extension, such as EXE, was limited to three characters and the period between them was probably not stored but artificially added so that "filename" and "exe" became filename.exe. The third component was a tad unusual as it consisted of a sequence number and may even have been stored in a few bits as a small integer, perhaps a byte and would be optionally shown with a semi-colon as a separator: filename.file-type;version DOS had an 8by3 version with no support for keeping track of sequence numbers. UNIX flowed the fields together with a limit of 14 that included an actual optional period as a counted character. Since then, many systems have allowed huge file names including with multiple periods as in open_me.tar.gz. So looking at the VMS model may not be as helpful for trying to come up with a way for a Python program to maintain a sequence of file names. VMS support for version numbers was coded widely within the OS and applications. I regularly had to purge and keep only recent versions because I kept running out of disk space. Many current OS do not have any such support and saving a file simply overwrites the earlier version by default. Some questions: What would a proposed scheme do with a filename that already ended legitimately with a number such as the address "broad23"? Would it save another version as broad24 or as broad231 or append more characters as in broad23-1? What do you do with file.txt as "file.txt1" would not work well or be recognized as a registered type? There are many more modern ways to manage versions of files and many of them involve placing full copies of the file somewhere else where they can be retrieved or making some kind of diff on the previous version that can allow reconstitution with some calculation using an original file and doing what modifications (additions, replacements, deletions) it takes to transition an original file to version1, then version 2 and so on. Other methods may involve squirrelling away complete copies along with metadata that lets them be searched and retrieved. Using just a filename change is probably easier for some things like a circular group of log files, though. FWIW, I see lots of computer languages (including Python) implementing what I consider typographical techniques where you implement a feature by taking a variable name and adding changes to get a result, such as is done with the name mangling feature where __NAME (double underscore) actually gets mapped to _someclass__NAME in the actual code. I see similar gimmicks in many languages like SCALA and RUST. This is sometimes a reasonable way to take a single string and add functionality with some manipulation -- but with some care as a user choosing odd names may cause unusual problems in some cases. -- https://mail.python.org/mailman/listinfo/python-list
