Issue 3028: tokenize module: normal lines, not "logical" (original) (raw)

Issue3028

Created on 2008-06-02 11:33 by noam, last changed 2022-04-11 14:56 by admin. This issue is now closed.

Messages (4)
msg67635 - (view)	Author: Noam Yorav-Raphael (noam)	Date: 2008-06-02 11:33
Hello, The documentation of the tokenize module says: "The line passed is the *logical* line; continuation lines are included." Some background: The tokenize module splits a python source into tokens, and says for each token where it begins and where it ends, in the format of (row, offset). This note in the documentation made me think that continuation lines are considered as one line, and made me break my head how I should find the offset of the token in the original string. The truth is that the row number is simply the index of the line as returned by the readline function, and it's very simple to reconstruct the string offset. I suggest that this will be changed to something like "The line passed is the index of the string returned by the readline function, plus 1. That is, the first string returned is called line 1, the second is called line 2, and so on." Thanks, Noam
msg67660 - (view)	Author: Senthil Kumaran (orsenthil) *	Date: 2008-06-03 17:14
> The documentation of the tokenize module says: "The line passed is the > *logical* line; continuation lines are included." > > I suggest that this will be changed to something like "The line passed > is the index of the string returned by the readline function, plus 1. > That is, the first string returned is called line 1, the second is > called line 2, and so on." The emphasis of *logical* may help us understand that it is a complete line. I find the wording of solution bit awkward, tough I am able to get what it is trying to say. - Index of string returned by readline function ?? and plus 1. ?? How about, The line passed is the *logical* non-blank line; continuation lines are included. The row counting starts from one.
msg67829 - (view)	Author: Georg Brandl (georg.brandl) *	Date: 2008-06-08 09:00
I got it wrong too at first; the "logical line" refers to the last tuple item, the line string, not the line *number* passed. It should now be clear as of r64037.
msg67832 - (view)	Author: Noam Yorav-Raphael (noam)	Date: 2008-06-08 14:47
Can I suggest that you also add something like "The row indices in the (row, column) tuples, however, are physical, and don't treat continuation lines specially."? It's just that it took me some time to understand your clarification, since the row indices I thought the documentation talks about are also tuple items, they just happen to be the first in the tuple, not the last.