[Patch v1.1] lib: add _notmuch_message_id_parse_strict

[David Bremner <david@tethera.net>]

The idea is that if a message-id parses with this function, the MUA
generating it was probably sane, and in particular it's probably safe
to use the result as a parent from In-Reply-to.
---

replace ^M with ^L in tests, hopefully less vulnerable to well meaning
MTAs

 lib/message-id.c        | 32 ++++++++++++++++++
 lib/notmuch-private.h   | 14 ++++++++
 test/Makefile.local     |  6 +++-
 test/T710-message-id.sh | 73 +++++++++++++++++++++++++++++++++++++++++
 test/message-id-parse.c | 24 ++++++++++++++
 5 files changed, 148 insertions(+), 1 deletion(-)
 create mode 100755 test/T710-message-id.sh
 create mode 100644 test/message-id-parse.c

diff --git a/lib/message-id.c b/lib/message-id.c
index d7541d50..a1dce9c8 100644
--- a/lib/message-id.c
+++ b/lib/message-id.c
@@ -1,4 +1,5 @@
 #include "notmuch-private.h"
+#include "string-util.h"
 
 /* Advance 'str' past any whitespace or RFC 822 comments. A comment is
  * a (potentially nested) parenthesized sequence with '\' used to
@@ -94,3 +95,34 @@ _notmuch_message_id_parse (void *ctx, const char *message_id, const char **next)
 
     return result;
 }
+
+char *
+_notmuch_message_id_parse_strict (void *ctx, const char *message_id)
+{
+    const char *s, *end;
+    char *result;
+
+    if (message_id == NULL || *message_id == '\0')
+	return NULL;
+
+    s = skip_space (message_id);
+    if (*s == '<')
+	s++;
+    else
+	return NULL;
+
+    for (end = s; *end && *end != '>'; end++)
+	if (isspace (*end))
+	    return NULL;
+
+    if (*end != '>')
+	return NULL;
+    else {
+	const char *last =  skip_space (end+1);
+	if (*last != '\0')
+	    return NULL;
+    }
+
+    result = talloc_strndup (ctx, s, end - s);
+    return result;
+}
diff --git a/lib/notmuch-private.h b/lib/notmuch-private.h
index 083fc013..67fd4990 100644
--- a/lib/notmuch-private.h
+++ b/lib/notmuch-private.h
@@ -526,6 +526,20 @@ _notmuch_query_count_documents (notmuch_query_t *query,
 char *
 _notmuch_message_id_parse (void *ctx, const char *message_id, const char **next);
 
+/* Parse a message-id, discarding leading and trailing whitespace, and
+ * '<' and '>' delimiters.
+ *
+ * Apply a probably-stricter-than RFC definition of what is allowed in
+ * a message-id. In particular, forbid whitespace.
+ *
+ * Returns a newly talloc'ed string belonging to 'ctx'.
+ *
+ * Returns NULL if there is any error parsing the message-id.
+ */
+
+char *
+_notmuch_message_id_parse_strict (void *ctx, const char *message_id);
+
 
 /* message.cc */
 
diff --git a/test/Makefile.local b/test/Makefile.local
index 1a0ab813..1cf09778 100644
--- a/test/Makefile.local
+++ b/test/Makefile.local
@@ -15,6 +15,9 @@ smtp_dummy_modules = $(smtp_dummy_srcs:.c=.o)
 $(dir)/arg-test: $(dir)/arg-test.o command-line-arguments.o util/libnotmuch_util.a
 	$(call quiet,CC) $^ -o $@ $(LDFLAGS)
 
+$(dir)/message-id-parse: $(dir)/message-id-parse.o lib/libnotmuch.a util/libnotmuch_util.a
+	$(call quiet,CC) $^ -o $@ $(LDFLAGS) $(TALLOC_LDFLAGS)
+
 $(dir)/hex-xcode: $(dir)/hex-xcode.o command-line-arguments.o util/libnotmuch_util.a
 	$(call quiet,CC) $^ -o $@ $(LDFLAGS) $(TALLOC_LDFLAGS)
 
@@ -50,7 +53,8 @@ test_main_srcs=$(dir)/arg-test.c \
 	      $(dir)/smtp-dummy.c \
 	      $(dir)/symbol-test.cc \
 	      $(dir)/make-db-version.cc \
-	      $(dir)/ghost-report.cc
+	      $(dir)/ghost-report.cc \
+	      $(dir)/message-id-parse.c
 
 test_srcs=$(test_main_srcs) $(dir)/database-test.c
 
diff --git a/test/T710-message-id.sh b/test/T710-message-id.sh
new file mode 100755
index 00000000..e73d6ba9
--- /dev/null
+++ b/test/T710-message-id.sh
@@ -0,0 +1,73 @@
+#!/usr/bin/env bash
+test_description="message id parsing"
+
+. $(dirname "$0")/test-lib.sh || exit 1
+
+test_begin_subtest "good message ids"
+${TEST_DIRECTORY}/message-id-parse <<EOF >OUTPUT
+<018b1a8f2d1df62e804ce88b65401304832dfbbf.1346614915.git.jani@nikula.org>
+<1530507300.raoomurnbf.astroid@strange.none>
+<1258787708-21121-2-git-send-email-keithp@keithp.com>
+EOF
+cat <<EOF >EXPECTED
+GOOD: 018b1a8f2d1df62e804ce88b65401304832dfbbf.1346614915.git.jani@nikula.org
+GOOD: 1530507300.raoomurnbf.astroid@strange.none
+GOOD: 1258787708-21121-2-git-send-email-keithp@keithp.com
+EOF
+test_expect_equal_file EXPECTED OUTPUT
+
+test_begin_subtest "leading and trailing space is OK"
+${TEST_DIRECTORY}/message-id-parse <<EOF >OUTPUT
+   <018b1a8f2d1df62e804ce88b65401304832dfbbf.1346614915.git.jani@nikula.org>
+<1530507300.raoomurnbf.astroid@strange.none>    
+    <1258787708-21121-2-git-send-email-keithp@keithp.com>
+EOF
+cat <<EOF >EXPECTED
+GOOD: 018b1a8f2d1df62e804ce88b65401304832dfbbf.1346614915.git.jani@nikula.org
+GOOD: 1530507300.raoomurnbf.astroid@strange.none
+GOOD: 1258787708-21121-2-git-send-email-keithp@keithp.com
+EOF
+test_expect_equal_file EXPECTED OUTPUT
+
+test_begin_subtest "<> delimeters are required"
+${TEST_DIRECTORY}/message-id-parse <<EOF >OUTPUT
+018b1a8f2d1df62e804ce88b65401304832dfbbf.1346614915.git.jani@nikula.org>
+<1530507300.raoomurnbf.astroid@strange.none
+1258787708-21121-2-git-send-email-keithp@keithp.com
+EOF
+cat <<EOF >EXPECTED
+BAD: 018b1a8f2d1df62e804ce88b65401304832dfbbf.1346614915.git.jani@nikula.org>
+BAD: <1530507300.raoomurnbf.astroid@strange.none
+BAD: 1258787708-21121-2-git-send-email-keithp@keithp.com
+EOF
+test_expect_equal_file EXPECTED OUTPUT
+
+test_begin_subtest "embedded whitespace is forbidden"
+${TEST_DIRECTORY}/message-id-parse <<EOF >OUTPUT
+<018b1a8f2d1df62e804ce88b65401304832dfbbf.1346614915 .git.jani@nikula.org>
+<1530507300.raoomurnbf.astroid	@strange.none>
+<1258787708-21121-\f2-git-send-email-keithp@keithp.com>
+EOF
+cat <<EOF >EXPECTED
+BAD: <018b1a8f2d1df62e804ce88b65401304832dfbbf.1346614915 .git.jani@nikula.org>
+BAD: <1530507300.raoomurnbf.astroid	@strange.none>
+BAD: <1258787708-21121-\f2-git-send-email-keithp@keithp.com>
+EOF
+test_expect_equal_file EXPECTED OUTPUT
+
+
+test_begin_subtest "folded real life bad In-Reply-To values"
+${TEST_DIRECTORY}/message-id-parse <<EOF >OUTPUT
+<22597.31869.380767.339702@chiark.greenend.org.uk> (Ian Jackson's message of "Mon, 5 Dec 2016 14:41:01 +0000")
+<20170625141242.loaalhis2eodo66n@gaara.hadrons.org>  <149719990964.27883.13021127452105787770.reportbug@seneca.home.org>
+Your message of Tue, 09 Dec 2014 13:21:11 +0100. <1900758.CgLNVPbY9N@liber>
+EOF
+cat <<EOF >EXPECTED
+BAD: <22597.31869.380767.339702@chiark.greenend.org.uk> (Ian Jackson's message of "Mon, 5 Dec 2016 14:41:01 +0000")
+BAD: <20170625141242.loaalhis2eodo66n@gaara.hadrons.org>  <149719990964.27883.13021127452105787770.reportbug@seneca.home.org>
+BAD: Your message of Tue, 09 Dec 2014 13:21:11 +0100. <1900758.CgLNVPbY9N@liber>
+EOF
+test_expect_equal_file EXPECTED OUTPUT
+
+
+test_done
diff --git a/test/message-id-parse.c b/test/message-id-parse.c
new file mode 100644
index 00000000..25decc9b
--- /dev/null
+++ b/test/message-id-parse.c
@@ -0,0 +1,24 @@
+#include <stdio.h>
+#include <talloc.h>
+#include "notmuch-private.h"
+
+int main(unused (int argc), unused (char **argv)){
+    char *line = NULL;
+    size_t len = 0;
+    ssize_t nread;
+    void *local = talloc_new (NULL);
+
+    while ((nread = getline(&line, &len, stdin)) != -1) {
+	int last = strlen(line) - 1;
+	if (line[last] == '\n')
+	    line[last] = '\0';
+
+	char *mid = _notmuch_message_id_parse_strict (local, line);
+	if (mid)
+	    printf("GOOD: %s\n", mid);
+	else
+	    printf("BAD: %s\n", line);
+    }
+
+    talloc_free (local);
+}
-- 
2.18.0