Remove "appendix" information from commit message

...similar to `git am`.

Add a new field, `BodyAppendix` to PatchHeader.

Modify `scanMessageBody` to return both the body and the appendix.
Do this by keeping two string builders, and having it switch to the
appendix builder when it finds a `---` line.

Handling the newlines at the end as expected requires moving things
around a bit.

First, we were trimming space from the line once to decide whether the
line was empty, and then trimming space again if we determined it
wasn't empty.  This only needs to be done once.

Then, do all the trimming (both of whitespace and the prefix) first,
before deciding what to do about the line.

Add some tests to verify that it works as expected.

NB that this patch will separate out an "appendix" even from the
output of `git log`, which presumably has already been checked in.  If
this is not the desired behavior, we could either:

1. Make the `---` check after trimming whitespace, but before trimming
the indent, or

2. Pass in a boolean to tell scanMessageBody not to separate out
appendix material if we're calling from parseHeaderPretty.

Signed-off-by: George Dunlap <[email protected]>

Author: George Dunlap

gitdiff/patch_header.go

@@ -44,6 +44,10 @@ type PatchHeader struct {
 	// remove prefixes such as `Re: ` and `[PATCH v3 5/17]` from the
 	// Title and place them here.
 	SubjectPrefix string
+
+	// If the preamble contains a `---` line, that line and everything
+	// after will be placed in BodyAppendix.
+	BodyAppendix string
 }
 
 // Message returns the commit message for the header. The message consists of
@@ -170,9 +174,17 @@ func ParsePatchDate(s string) (time.Time, error) {
 // `[PATCH]` or `Re:` in the same way that `git mailinfo` does.
 // SubjectPrefix will be set to the value of this removed string.
 // (`git mailinfo` is the core part of `git am` that pulls information
-// out of an individual mail.)  Unline `git mailinfo`,
-// ParsePatchHeader does not at the moment remove commit states or
-// other extraneous matter after a `---` line.
+// out of an individual mail.)
+//
+// Additionally, ParsePatchHeader will separate a `---` line and
+// anything after it into BodyAppendix.
+//
+// Those wishing the effect of a plain `git am` should use
+// `PatchHeader.Title + "\n" + PatchHeader.Body` (or
+// `PatchHeader.Message()`).  Those wishing to retain the subject
+// prefix and appendix material should use `PatchHeader.SubjectPrefix
+// + PatchHeader.Title + "\n" + PatchHeader.Body + "\n" +
+// PatchHeader.BodyAppendix`.
 func ParsePatchHeader(s string) (*PatchHeader, error) {
 	r := bufio.NewReader(strings.NewReader(s))
 
@@ -277,11 +289,12 @@ func parseHeaderPretty(prettyLine string, r io.Reader) (*PatchHeader, error) {
 	h.Title = title
 
 	if title != "" {
-		body := scanMessageBody(s, indent)
+		body, appendix := scanMessageBody(s, indent)
 		if s.Err() != nil {
 			return nil, s.Err()
 		}
 		h.Body = body
+		h.BodyAppendix = appendix
 	}
 
 	return h, nil
@@ -309,29 +322,39 @@ func scanMessageTitle(s *bufio.Scanner) (title string, indent string) {
 	return b.String(), indent
 }
 
-func scanMessageBody(s *bufio.Scanner, indent string) string {
-	var b strings.Builder
+func scanMessageBody(s *bufio.Scanner, indent string) (string, string) {
+	// Body and appendix
+	var body, appendix strings.Builder
+	c := &body
 	var empty int
 	for i := 0; s.Scan(); i++ {
 		line := s.Text()
-		if strings.TrimSpace(line) == "" {
+
+		line = strings.TrimRightFunc(line, unicode.IsSpace)
+		line = strings.TrimPrefix(line, indent)
+
+		if line == "" {
 			empty++
 			continue
 		}
 
-		if b.Len() > 0 {
-			b.WriteByte('\n')
+		// Parse out "appendix" information (often added by `git
+		// format-patch` and removed by `git am`).
+		if c == &body && line == "---" {
+			c = &appendix
+		}
+
+		if c.Len() > 0 {
+			c.WriteByte('\n')
 			if empty > 0 {
-				b.WriteByte('\n')
+				c.WriteByte('\n')
 			}
 		}
 		empty = 0
 
-		line = strings.TrimRightFunc(line, unicode.IsSpace)
-		line = strings.TrimPrefix(line, indent)
-		b.WriteString(line)
+		c.WriteString(line)
 	}
-	return b.String()
+	return body.String(), appendix.String()
 }
 
 func parseHeaderMail(mailLine string, r io.Reader) (*PatchHeader, error) {
@@ -372,7 +395,7 @@ func parseHeaderMail(mailLine string, r io.Reader) (*PatchHeader, error) {
 	h.SubjectPrefix, h.Title = parseSubject(subject)
 
 	s := bufio.NewScanner(msg.Body)
-	h.Body = scanMessageBody(s, "")
+	h.Body, h.BodyAppendix = scanMessageBody(s, "")
 	if s.Err() != nil {
 		return nil, s.Err()
 	}

gitdiff/patch_header_test.go

@@ -139,6 +139,7 @@ func TestParsePatchHeader(t *testing.T) {
 	expectedDate := time.Date(2020, 04, 11, 15, 21, 23, 0, time.FixedZone("PDT", -7*60*60))
 	expectedTitle := "A sample commit to test header parsing"
 	expectedBody := "The medium format shows the body, which\nmay wrap on to multiple lines.\n\nAnother body line."
+	expectedBodyAppendix := "---\nCC: Joe Smith <[email protected]>"
 
 	tests := map[string]struct {
 		Input  string
@@ -221,6 +222,33 @@ CommitDate: Sat Apr 11 15:21:23 2020 -0700
 				Body:          expectedBody,
 			},
 		},
+		"prettyAppendix": {
+			Input: `commit 61f5cd90bed4d204ee3feb3aa41ee91d4734855b
+Author:     Morton Haypenny <[email protected]>
+AuthorDate: Sat Apr 11 15:21:23 2020 -0700
+Commit:     Morton Haypenny <[email protected]>
+CommitDate: Sat Apr 11 15:21:23 2020 -0700
+
+    A sample commit to test header parsing
+
+    The medium format shows the body, which
+    may wrap on to multiple lines.
+
+    Another body line.
+    ---
+    CC: Joe Smith <[email protected]>
+`,
+			Header: PatchHeader{
+				SHA:           expectedSHA,
+				Author:        expectedIdentity,
+				AuthorDate:    expectedDate,
+				Committer:     expectedIdentity,
+				CommitterDate: expectedDate,
+				Title:         expectedTitle,
+				Body:          expectedBody,
+				BodyAppendix:  expectedBodyAppendix,
+			},
+		},
 		"mailbox": {
 			Input: `From 61f5cd90bed4d204ee3feb3aa41ee91d4734855b Mon Sep 17 00:00:00 2001
 From: Morton Haypenny <[email protected]>
@@ -240,6 +268,28 @@ Another body line.
 				Body:       expectedBody,
 			},
 		},
+		"mailboxAppendix": {
+			Input: `From 61f5cd90bed4d204ee3feb3aa41ee91d4734855b Mon Sep 17 00:00:00 2001
+From: Morton Haypenny <[email protected]>
+Date: Sat, 11 Apr 2020 15:21:23 -0700
+Subject: [PATCH] A sample commit to test header parsing
+
+The medium format shows the body, which
+may wrap on to multiple lines.
+
+Another body line.
+---
+CC: Joe Smith <[email protected]>
+`,
+			Header: PatchHeader{
+				SHA:          expectedSHA,
+				Author:       expectedIdentity,
+				AuthorDate:   expectedDate,
+				Title:        expectedTitle,
+				Body:         expectedBody,
+				BodyAppendix: expectedBodyAppendix,
+			},
+		},
 		"unwrapTitle": {
 			Input: `commit 61f5cd90bed4d204ee3feb3aa41ee91d4734855b
 Author: Morton Haypenny <[email protected]>
@@ -333,6 +383,10 @@ Author: Morton Haypenny <[email protected]>
 			if exp.Body != act.Body {
 				t.Errorf("incorrect parsed body:\n  expected: %q\n    actual: %q", exp.Body, act.Body)
 			}
+			if exp.BodyAppendix != act.BodyAppendix {
+				t.Errorf("incorrect parsed body appendix:\n  expected: %q\n    actual: %q",
+					exp.BodyAppendix, act.BodyAppendix)
+			}
 		})
 	}
 }