Replies: 1 comment
|
there's a partial version of this already, which you may not have run into: if your annotation text contains the word "hunk" (whole word, case-insensitive), the exported header auto-expands from the single line to the full contiguous change range, e.g. beyond that, my take is that a single-line anchor already carries the range implicitly. The exported annotation includes the file, the line, and the surrounding diff context, and the comment text usually makes the scope obvious. An AI agent reading "lines 3-10 are one unit, refactor together" pinned to line 10 gets it without a formal range. So for most cases line + the hunk keyword covers it. the one thing neither covers cleanly is a deliberately non-contiguous or cross-hunk selection (say lines 3, 7 and 10 belong together), but prose handles that fine too, and a visual range-select mode is a fair bit of UI/state for a fairly niche win. does the hunk keyword cover your case? If not, can you share a concrete example where a single-line annotation plus the comment text genuinely loses information? Happy to reconsider if there's a real gap. |
Uh oh!
There was an error while loading. Please reload this page.
When adding an annotation, I want to signal to the reader that my annotation is not about a specific line in particular, but a group of lines combined.
E.g. lines 3-10 contain a unit of work that i want to comment on as a whole. Right now I add the annotation to line 10 and note that my comment spans line 3-10, it would be ideal if this could be codified in Revdiff.
All reactions