[0cf8612b2e1e]

If you can use a string anywhere you can use a t-string, then a non parametrized t-string is a code smell (lining error). If there is a dedicated template-string API, then there is the implicit threat you are breaking backwards compatibility to stop using regular strings.

[davepeck]

> If you can use a string anywhere you can use a t-string

You can't; they're different types. t-strings are not `str`

It's up to good framework/API design to take advantage of this.

[0cf8612b2e1e]

A library writer ultimately has to decide if they accept both types. For a database cursor, do you take regular strings + parameter arguments and template strings? Or dedicate a new API to the idea?

  cursor.execute(“select * from x where foo=?”, {foo=1})
  # while also allowing
  cursor.execute(t“select * from x where foo={foo}”)
  #Vs 
  cursor.executetemplate(“select * from x where foo={foo}”)

If ‘execute’ takes string and t-string, then I would consider it a problem to use a t-string without parameters. If there is a novel API just for t-strings, then you are implying widespread breaking changes as you have a schism between the two ways of providing parameters.

[davepeck]

My point was that library authors will need to consider this carefully. If you're writing a library where injection attacks matter, then -- long term -- you almost certainly do not want a single method that accepts `Union[str, Template]`. You probably either want to avoid accepting `str` entirely, or perhaps provide two separate methods. Some period of deprecation seems inevitable.

[WorldMaker]

Yeah, "t-string" is possibly a misnomer, because they are in fact at runtime a Template object (from string.templatelib).