colorize: emit OSC 8 for URLs and emails

Mark URLs with OSC 8 escape sequence to help terminal emulators with
opening multi-line URLs with the mouse and attach the hyperlink to
email addresses, so the users could open them.

Link: https://gist.github.com/egmontkob/eb114294efbcd5adb1944c9f3cb5feda
Signed-off-by: Kirill Chibisov <contact@kchibisov.com>
Acked-by: Robin Jarry <robin@jarry.cc>

3 files changed, 27 insertions, 7 deletions

diff --git a/filters/colorize.c b/filters/colorize.c
index e1d57be6..26c25a8a 100644
--- a/filters/colorize.c
+++ b/filters/colorize.c
@@ -413,6 +413,17 @@ static inline size_t print_notabs(const char *in, size_t max_len)
 	return len;
 }
 
+static void print_osc8(const char *url, size_t len, size_t id, bool email) {
+	printf("\x1b]8;id=colorize-%lu;", id);
+	if (url != NULL) {
+		if (email) {
+			print("mailto://");
+		}
+		print_notabs(url, len);
+	}
+	print("\x1b\\");
+}
+
 static void diff_chunk(const char *in)
 {
 	size_t len = 0;
@@ -429,17 +440,22 @@ static void diff_chunk(const char *in)
 }
 
 #define URL_RE \
-	"[a-z]{2,8}://[][:alnum:]._~:/?#[@!$&'()*+,;=%-]{4,}" \
+	"([a-z]{2,8}:)//[][:alnum:]._~:/?#[@!$&'()*+,;=%-]{4,}" \
 	"|(mailto:)?[[:alnum:]_+.~/-]*[[:alnum:]]@[a-z][[:alnum:].-]*[a-z]"
 static regex_t url_re;
 
 static void urls(const char *in, struct style *ctx)
 {
-	regmatch_t groups[2];
+	/* ID of the next link to print for OSC 8. The purpose of passing
+	 * explicit ID is to help terminal emulator with grouping of
+	 * multi-line links in nested terminal sessions */
+	static size_t url_id = 0;
+
+	regmatch_t groups[3];
 	size_t len;
 	bool trim;
 
-	while (!regexec(&url_re, in, 2, groups, 0)) {
+	while (!regexec(&url_re, in, 3, groups, 0)) {
 		in += print_notabs(in, groups[0].rm_so);
 		print(seq(&styles.url));
 		len = groups[0].rm_eo - groups[0].rm_so;
@@ -457,7 +473,11 @@ static void urls(const char *in, struct style *ctx)
 				break;
 			}
 		}
+		bool email = groups[2].rm_so == -1 && groups[1].rm_so == -1;
+		print_osc8(in, len, url_id, email);
 		in += print_notabs(in, len);
+		print_osc8(NULL, 0, url_id, email);
+		url_id++;
 		print(RESET);
 		if (ctx) {
 			print(seq(ctx));
diff --git a/filters/vectors/colorize-patch.expected b/filters/vectors/colorize-patch.expected
index de3f2399..619e118f 100644
--- a/filters/vectors/colorize-patch.expected
+++ b/filters/vectors/colorize-patch.expected
@@ -1,11 +1,11 @@
-From: Robin Jarry <robin@jarry.cc>
+From: Robin Jarry <]8;id=colorize-0;mailto://robin@jarry.cc\robin@jarry.cc]8;id=colorize-0;\>
 Date: Mon, 26 Dec 2022 17:02:14 +0100
 Subject: [PATCH aerc] doc: fix numbered lists
 
 According to scdoc(5), numbered lists start with a period.
 
 Fixes: af63bd0188d1 ("doc: homogenize scdoc markup")
-Signed-off-by: Robin Jarry <robin@jarry.cc>
+Signed-off-by: Robin Jarry <]8;id=colorize-1;mailto://robin@jarry.cc\robin@jarry.cc]8;id=colorize-1;\>
 ---
  doc/aerc-stylesets.7.scd | 18 +++++++++---------
  1 file changed, 9 insertions(+), 9 deletions(-)
diff --git a/filters/vectors/colorize-quotes.expected b/filters/vectors/colorize-quotes.expected
index 6e71d60a..46d3adba 100644
--- a/filters/vectors/colorize-quotes.expected
+++ b/filters/vectors/colorize-quotes.expected
@@ -17,7 +17,7 @@ nibh petentium at! Sit docendi laboramus ei, animal insolens ad mea.
 > Erat alienum interpretaris has et, te vim aliquam molestie. Nam vivendum
 > facilisis qualisque at, ex his mucius qualisque! Fabulas lucilius adversarium
 > eu his. Cu soluta inermis accusata usu, his nulla dolore ne, vis id semper
-> detracto sententia <https://foobar.com> && "https://foobaz.org/".
+> detracto sententia <]8;id=colorize-0;https://foobar.com\https://foobar.com]8;id=colorize-0;\> && "]8;id=colorize-1;https://foobaz.org/\https://foobaz.org/]8;id=colorize-1;\".
 >
 > Error libris deleniti ea mei, vis at elit probo munere, his sint unum
 > albucius ex.
@@ -52,7 +52,7 @@ Has putent verterem constituto ex, tale electram duo at! Ei nulla lucilius
 intellegat nam, pro quod epicuri dissentiet ut, omnis voluptatibus definitiones
 vim at.
 
-https://git-man-page-generator.lokaltog.net/#Y2xhcCQkY29tbWFuZA==
+]8;id=colorize-2;https://git-man-page-generator.lokaltog.net/#Y2xhcCQkY29tbWFuZA==\https://git-man-page-generator.lokaltog.net/#Y2xhcCQkY29tbWFuZA==]8;id=colorize-2;\
 
 Eam mundi libris debitis ad, eam regione numquam at. Eum omnes bonorum eu,
 oporteat assueverit disputationi nam ne, nonumes iracundia mea ad! Duo libris