fix split to be html tag aware

[dselans]

Note

Medium Risk
Changes how outbound email HTML is segmented into Kustomer variables, which could affect rendering if edge cases (e.g., no > near the boundary) behave unexpectedly; coverage is improved via new tests.

Overview
Updates chunkHTMLBody to advance by variable chunk lengths and, when possible, cut each 1024-char chunk at the last > boundary so HTML tags aren’t broken across Kustomer custom variables.

Adds a focused test suite covering single-chunk behavior, tag-aware splitting at boundaries, reassembly correctness, and fallback behavior when no tags are present.

^{Written by Cursor Bugbot for commit a44272b. This will update automatically on new commits. Configure here.}

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

Bugbot Autofix prepared a fix for the issue found in the latest run.

• ✅ Fixed: Test asserts wrong suffix, missing actual mid-tag split
  • Fixed chunkHTMLBody to fall back to splitting before the last '<' when no '>' is found (preventing mid-tag splits), and fixed the test assertion to check that the first chunk equals the pre-tag content exactly instead of checking for the wrong partial suffix.

Or push these changes by commenting:

@cursor push 2acaf74167

Preview (2acaf74167)

diff --git a/kustomer.go b/kustomer.go
--- a/kustomer.go
+++ b/kustomer.go
@@ -728,6 +728,8 @@
 			lastClose := strings.LastIndex(input[start:end], ">")
 			if lastClose != -1 {
 				end = start + lastClose + 1
+			} else if lastOpen := strings.LastIndex(input[start:end], "<"); lastOpen > 0 {
+				end = start + lastOpen
 			}
 		}
 

diff --git a/kustomer_test.go b/kustomer_test.go
--- a/kustomer_test.go
+++ b/kustomer_test.go
@@ -227,7 +227,7 @@
 			Expect(reassembled).To(Equal(msg))
 
 			// First chunk must end at or before the '<' of the tag
-			Expect(strings.HasSuffix(chunks["htmlContent0Str"], "<stro")).To(BeFalse())
+			Expect(chunks["htmlContent0Str"]).To(Equal(before))
 		})
 
 		It("reassembles to the original content", func() {

[cursor]

Test asserts wrong suffix, missing actual mid-tag split

Medium Severity

The test "does not split in the middle of an HTML tag" doesn't actually verify its claim. With 1020 xs before <strong>, there's no > anywhere in input[0:1024], so lastClose is -1 and the function falls back to splitting at position 1024 — right in the middle of <strong>. The first chunk ends with <str, which IS a mid-tag split. But the assertion only checks that the suffix isn't <stro (off by one character), so the test passes despite the function doing exactly what it claims not to do.