feat(coverage): add changed-lines helper and diff parser

Second of three pure helpers for the coverage-report command.

cj/--coverage-parse-diff-output is pure. It takes a git unified-diff string and returns a hash table of file to set of added or modified line numbers (based on the +new_start,new_count hunk headers). Files with deletion-only hunks appear in the result with an empty set, so reporters can distinguish "coverage not tracked" from "no changes touched this file."

cj/--coverage-changed-lines wraps that parser with scope dispatch. Scopes are working-tree, staged, branch-vs-main, and branch-vs-parent. Branch-vs-parent takes an optional BASE arg; if omitted, falls back to @{upstream}. Unknown scopes signal user-error.

Tests cover Normal (single hunk, multiple files), Boundary (new file via @@ -0,0, deletion-only, binary markers, single-line hunks without a count, empty input), and Error (malformed hunk headers skipped; unknown scope). Git invocation is stubbed via cl-letf in the smoke test so the parser logic is exercised without shelling out.

Part of the coverage-core work per docs/design/coverage.org.

1 files changed, 64 insertions, 0 deletions

diff --git a/modules/coverage-core.el b/modules/coverage-core.el
index 4a3112ce..a2b03817 100644
--- a/modules/coverage-core.el
+++ b/modules/coverage-core.el
@@ -54,5 +54,69 @@ Signals `user-error' if FILE does not exist."
 		(forward-line 1)))
 	result))
 
+(defconst cj/--coverage-hunk-header-regexp
+  "^@@ -[0-9]+\\(,[0-9]+\\)? \\+\\([0-9]+\\)\\(,\\([0-9]+\\)\\)? @@"
+  "Regexp for a git unified-diff hunk header.
+Captures new_start (group 2) and new_count (group 4; nil implies 1).")
+
+(defconst cj/--coverage-file-marker-regexp
+  "^\\+\\+\\+ b/\\(.+\\)$"
+  "Regexp for the \"+++ b/<path>\" line of a git diff.
+Captures the file path (group 1).")
+
+(defun cj/--coverage-parse-diff-output (output)
+  "Parse OUTPUT, a git unified-diff string, into a hash table.
+Keys are file paths (relative to repo root, as git emits them).  Values
+are hash tables whose keys are line numbers added or modified in the new
+version of the file.  A file that appears with only deletions maps to an
+empty hash table.  Malformed hunk headers are skipped silently."
+  (let ((result (make-hash-table :test 'equal))
+		(current-lines nil))
+	(with-temp-buffer
+	  (insert output)
+	  (goto-char (point-min))
+	  (while (not (eobp))
+		(let ((line (buffer-substring-no-properties
+					 (line-beginning-position) (line-end-position))))
+		  (cond
+		   ((string-match cj/--coverage-file-marker-regexp line)
+			(let ((path (match-string 1 line)))
+			  (setq current-lines (make-hash-table :test 'eql))
+			  (puthash path current-lines result)))
+		   ((string-prefix-p "+++ /dev/null" line)
+			(setq current-lines nil))
+		   ((and current-lines
+				 (string-match cj/--coverage-hunk-header-regexp line))
+			(let* ((new-start (string-to-number (match-string 2 line)))
+				   (count-str (match-string 4 line))
+				   (new-count (if count-str
+								  (string-to-number count-str)
+								1)))
+			  (when (> new-count 0)
+				(dotimes (i new-count)
+				  (puthash (+ new-start i) t current-lines)))))))
+		(forward-line 1)))
+	result))
+
+(defun cj/--coverage-changed-lines (scope &optional base)
+  "Return a hash table of files to changed line numbers for SCOPE.
+SCOPE is one of the symbols `working-tree', `staged', `branch-vs-main',
+or `branch-vs-parent'.  For `branch-vs-parent', BASE is the ref to
+compare against; if nil, falls back to the tracked upstream @{upstream}.
+Signals `user-error' for any other SCOPE."
+  (let ((cmd (cond
+			  ((eq scope 'working-tree)
+			   "git diff HEAD --unified=0")
+			  ((eq scope 'staged)
+			   "git diff --cached --unified=0")
+			  ((eq scope 'branch-vs-main)
+			   "git diff $(git merge-base HEAD main)..HEAD --unified=0")
+			  ((eq scope 'branch-vs-parent)
+			   (format "git diff $(git merge-base HEAD %s)..HEAD --unified=0"
+					   (or base "@{upstream}")))
+			  (t
+			   (user-error "Unknown coverage scope: %s" scope)))))
+	(cj/--coverage-parse-diff-output (shell-command-to-string cmd))))
+
 (provide 'coverage-core)
 ;;; coverage-core.el ends here