filters: rewrite colorize in c

Since its introduction, we had multiple issues with the colorize awk
script with regard to non-GNU awk compatibility.

Also, this script is standalone and the color theme must be hard coded
into it. Reading from an external configuration file (aerc's styleset)
from a non-GNU awk is close to impossible (and even far from trivial
with GNU awk).

Rewrite the builtin colorize filter in C to allow getting the color
theme from aerc's active styleset. The theme is configured using the
existing styleset syntax and attributes under a separate [viewer]
section (see examples and man page).

Export the active styleset file path to AERC_STYLESET env var when
invoking the filter command so that colorize can access it and use it.

I have tested compilation (with clang-analyzer and gcc -fanalyzer) and
basic operation on FreeBSD, Fedora (glibc) and Alpine (muslibc). More
tests would probably be required on MacOSX and older Linux distros.
I also added test vectors to give some confidence that this works as
expected. The execution with these vectors passed valgrind
--leak-check=full without errors.

NB: the default theme has changed to be more minimal. Sample stylesets
have more colorful examples. The awk -v theme=xxx option is no longer
supported.

  usage: colorize [-h] [-s FILE] [-f FILE]

  options:
    -h       show this help message
    -s FILE  use styleset file (default $AERC_STYLESET)
    -f FILE  read from filename (default stdin)

Signed-off-by: Robin Jarry <robin@jarry.cc>
Tested-by: Bence Ferdinandy <bence@ferdinandy.com>
Acked-by: Moritz Poldrack <moritz@poldrack.dev>

6 files changed, 909 insertions, 177 deletions

diff --git a/filters/colorize b/filters/colorize
deleted file mode 100755
index 9d2a7363..00000000
--- a/filters/colorize
+++ /dev/null
@@ -1,177 +0,0 @@
-#!/usr/bin/awk -f
-# Copyright (c) 2022 Robin Jarry
-#
-# A filter for the aerc mail program to colorize messages / attachments.
-# Basic colour themes are supported. To use a theme set the theme variable
-# in your aerc.conf accordingly, for example:
-#
-# text/plain=colorize -v theme=solarized
-
-BEGIN {
-	if (theme == "solarized") {
-		# R;G;B colors
-		url = "\033[38;2;181;137;0m" # yellow
-		header = "\033[38;2;211;54;130m" # magenta
-		signature = "\033[38;2;211;54;130m" # magenta
-		diff_meta = "\033[1;38;2;131;148;150m" # bold brblue
-		diff_chunk = "\033[38;2;42;161;152m" # cyan
-		diff_add = "\033[38;2;133;153;0m" # green
-		diff_del = "\033[38;2;220;50;47m" # red
-		quote_1 = "\033[38;2;38;139;210m"  # blue
-		quote_2 = "\033[38;2;203;75;22m" # brred
-		quote_3 = "\033[38;2;211;54;130m" # magenta
-		quote_4 = "\033[38;2;108;113;196m" # brmagenta
-		quote_x = "\033[38;2;147;161;161m" # brcyan
-		bold = "\033[1m"
-		reset = "\033[0m"
-	} else if (theme == "" || theme == "default") {
-		# R;G;B colors
-		url = "\033[38;2;255;255;175m" # yellow
-		header = "\033[38;2;175;135;255m" # purple
-		signature = "\033[38;2;175;135;255m" # purple
-		diff_meta = "\033[1;38;2;255;255;255m" # bold white
-		diff_chunk = "\033[38;2;0;205;205m" # cyan
-		diff_add = "\033[38;2;0;205;0m" # green
-		diff_del = "\033[38;2;205;0;0m" # red
-		quote_1 = "\033[38;2;95;175;255m"  # blue
-		quote_2 = "\033[38;2;255;135;0m" # orange
-		quote_3 = "\033[38;2;175;135;255m" # purple
-		quote_4 = "\033[38;2;255;95;215m" # pink
-		quote_x = "\033[38;2;128;128;128m" # gray
-		bold = "\033[1m"
-		reset = "\033[0m"
-	} else if (theme == "terminal") {
-		# terminal respects the users configured terminal color theme
-		url = "\033[4;34m" # underline blue
-		header = "\033[35m" # magenta
-		signature = "\033[35m" # magenta
-		diff_meta = "\033[2m" # faint
-		diff_chunk = "\033[36m" # cyan
-		diff_add = "\033[32m" # green
-		diff_del = "\033[31m" # red
-		quote_1 = "\033[37m"  # grey
-		quote_2 = "\033[34m" # blue
-		quote_3 = "\033[2;37m" # faint grey
-		quote_4 = "\033[2;34m" # faint blue
-		quote_x = "\033[2;37m" # faint grey
-		bold = "\033[1m"
-		reset = "\033[0m"
-	} else {
-		print "error: unknown theme " theme > "/dev/stderr"
-		exit 1
-	}
-	# state
-	in_diff = 0
-	in_signature = 0
-	in_headers = 0
-	in_body = 0
-	# patterns
-	header_pattern = "^[A-Z][[:alnum:]-]+:"
-	url_pattern = "[[:lower:]]+://[[:graph:]]+|(mailto:)?[[:alnum:]_\\+\\.~/-]*[[:alnum:]_]@[[:lower:]][[:alnum:]\\.-]*[[:lower:]]"
-	meta_pattern = "^(diff --git|(new|deleted) file|similarity index|(rename|copy) (to|from)|index|---|\\+\\+\\+) "
-}
-function color_quote(line) {
-	level = 0
-	quotes = ""
-	while (line ~ /^>/) {
-		level += 1
-		quotes = quotes ">"
-		line = substr(line, 2)
-		while (line ~ /^ /) {
-			quotes = quotes " "
-			line = substr(line, 2)
-		}
-	}
-	if (level == 1) {
-		color = quote_1
-	} else if (level == 2) {
-		color = quote_2
-	} else if (level == 3) {
-		color = quote_3
-	} else if (level == 4) {
-		color = quote_4
-	} else {
-		color = quote_x
-	}
-	if (match(line, meta_pattern)) {
-		return color quotes bold line reset
-	} else if (line ~ /^\+/) {
-		return color quotes diff_add line reset
-	} else if (line ~ /^-/) {
-		return color quotes diff_del line reset
-	}
-	gsub(url_pattern, url "&" color, line)
-	return color quotes line reset
-}
-{
-	# Strip carriage returns from line
-	sub(/\r$/, "")
-
-	if (in_diff) {
-		if ($0 ~ /^-- ?$/) {
-			in_diff = 0
-			in_signature = 1
-			$0 = signature $0 reset
-		} else if ($0 ~ /^@@ /) {
-			gsub(/^@@[^@]+@@/, diff_chunk "&" reset)
-		} else if (match($0, meta_pattern)) {
-			$0 = diff_meta $0 reset
-		} else if ($0 ~ /^\+/) {
-			$0 = diff_add $0 reset
-		} else if ($0 ~ /^-/) {
-			$0 = diff_del $0 reset
-		} else if ($0 !~ /^ / && $0 !~ /^$/) {
-			in_diff = 0
-			in_body = 1
-			if ($0 ~ /^>/) {
-				$0 = color_quote($0)
-			} else {
-				gsub(url_pattern, url "&" reset)
-			}
-		}
-	} else if (in_signature) {
-		gsub(url_pattern, url "&" signature)
-		$0 = signature $0 reset
-	} else if (in_headers) {
-		if ($0 ~ /^$/) {
-			in_headers = 0
-			in_body = 1
-		} else {
-			sub(header_pattern, header "&" reset)
-			gsub(url_pattern, url "&" reset)
-		}
-	} else if (in_body) {
-		if ($0 ~ /^>/) {
-			$0 = color_quote($0)
-		} else if ($0 ~ /^diff --git /) {
-			in_body = 0
-			in_diff = 1
-			$0 = diff_meta $0 reset
-		} else if ($0 ~ /^-- ?$/) {
-			in_body = 0
-			in_signature = 1
-			$0 = signature $0 reset
-		} else {
-			gsub(url_pattern, url "&" reset)
-		}
-	} else if ($0 ~ /^diff --git /) {
-		in_diff = 1
-		$0 = diff_meta $0 reset
-	} else if ($0 ~ /^-- ?$/) {
-		in_signature = 1
-		$0 = signature $0 reset
-	} else if (match($0, header_pattern)) {
-		in_headers = 1
-		sub(header_pattern, header "&" reset)
-		gsub(url_pattern, url "&" reset)
-	} else {
-		in_body = 1
-		if ($0 ~ /^>/) {
-			$0 = color_quote($0)
-		} else {
-			gsub(url_pattern, url "&" reset)
-		}
-	}
-
-	print
-}
diff --git a/filters/colorize.c b/filters/colorize.c
new file mode 100644
index 00000000..17eb548a
--- /dev/null
+++ b/filters/colorize.c
@@ -0,0 +1,681 @@
+/* SPDX-License-Identifier: MIT */
+/* Copyright (c) 2023 Robin Jarry */
+
+#include <errno.h>
+#include <fnmatch.h>
+#include <getopt.h>
+#include <regex.h>
+#include <stdint.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+static void usage(void)
+{
+	puts("usage: colorize [-h] [-s FILE] [-f FILE]");
+	puts("");
+	puts("options:");
+	puts("  -h       show this help message");
+	puts("  -s FILE  use styleset file (default $AERC_STYLESET)");
+	puts("  -f FILE  read from filename (default stdin)");
+}
+
+enum color_type {
+	NONE = 0,
+	DEFAULT,
+	RGB,
+	PALETTE,
+};
+
+struct color {
+	enum color_type type;
+	uint32_t rgb;
+	uint32_t index;
+};
+
+struct style {
+	struct color fg;
+	struct color bg;
+	int bold;
+	int blink;
+	int underline;
+	int reverse;
+	int italic;
+	int dim;
+	char *sequence;
+};
+
+__attribute__((malloc,returns_nonnull))
+static void *xmalloc(size_t s)
+{
+	void *ptr = malloc(s);
+	if (ptr == NULL) {
+		perror("fatal: cannot allocate buffer");
+		abort();
+	}
+	return ptr;
+}
+
+#define BOLD "\x1b[1m"
+#define RESET "\x1b[0m"
+#define LONGEST_SEQ "\x1b[1;2;3;4;5;7;38;2;255;255;255;48;2;255;255;255m"
+
+const char *seq(struct style *s) {
+	if (!s->sequence) {
+		char *b, *buf = xmalloc(strlen(LONGEST_SEQ) + 1);
+		const char *sep = "";
+
+		b = buf;
+		b += sprintf(b, "%s", "\x1b[");
+		if (s->bold) {
+			b += sprintf(b, "%s1", sep);
+			sep = ";";
+		}
+		if (s->dim) {
+			b += sprintf(b, "%s2", sep);
+			sep = ";";
+		}
+		if (s->italic) {
+			b += sprintf(b, "%s3", sep);
+			sep = ";";
+		}
+		if (s->underline) {
+			b += sprintf(b, "%s4", sep);
+			sep = ";";
+		}
+		if (s->blink) {
+			b += sprintf(b, "%s5", sep);
+			sep = ";";
+		}
+		if (s->reverse) {
+			b += sprintf(b, "%s7", sep);
+			sep = ";";
+		}
+		switch (s->fg.type) {
+		case NONE:
+			break;
+		case DEFAULT:
+			b += sprintf(b, "%s39", sep);
+			break;
+		case RGB:
+			b += sprintf(b, "%s38;2;%d;%d;%d", sep,
+				(s->fg.rgb >> 16) & 0xff,
+				(s->fg.rgb >> 8) & 0xff,
+				s->fg.rgb & 0xff);
+			sep = ";";
+			break;
+		case PALETTE:
+			b += sprintf(b, (s->fg.index < 8) ?
+				"%s3%d" : "%s38;5;%d", sep, s->fg.index);
+			sep = ";";
+			break;
+		}
+		switch (s->bg.type) {
+		case NONE:
+			break;
+		case DEFAULT:
+			b += sprintf(b, "%s49", sep);
+			break;
+		case RGB:
+			b += sprintf(b, "%s48;2;%d;%d;%d", sep,
+				(s->bg.rgb >> 16) & 0xff,
+				(s->bg.rgb >> 8) & 0xff,
+				s->bg.rgb & 0xff);
+			break;
+		case PALETTE:
+			b += sprintf(b, (s->bg.index < 8) ?
+				"%s4%d" : "%s48;5;%d", sep, s->bg.index);
+			break;
+		}
+		if (strcmp(buf, "\x1b[") == 0) {
+			b += sprintf(b, "0");
+		}
+		sprintf(b, "m");
+		s->sequence = buf;
+	}
+	return s->sequence;
+}
+
+struct styles {
+	struct style url;
+	struct style header;
+	struct style signature;
+	struct style diff_meta;
+	struct style diff_chunk;
+	struct style diff_add;
+	struct style diff_del;
+	struct style quote_1;
+	struct style quote_2;
+	struct style quote_3;
+	struct style quote_4;
+	struct style quote_x;
+};
+
+static FILE *in_file;
+static const char *styleset;
+static struct styles styles = {
+	.url = { .underline = 1 },
+	.header = { .bold = 1 },
+	.signature = { .dim = 1 },
+	.diff_meta = { .bold = 1 },
+	.diff_chunk = { .dim = 1 },
+	.diff_add = { .fg = { .type = PALETTE, .index = 2 } },
+	.diff_del = { .fg = { .type = PALETTE, .index = 1 } },
+	.quote_1 = { .fg = { .type = PALETTE, .index = 6 } },
+	.quote_2 = { .fg = { .type = PALETTE, .index = 6 }, .dim = 1 },
+	.quote_3 = { .fg = { .type = PALETTE, .index = 6 }, .dim = 1 },
+	.quote_4 = { .fg = { .type = PALETTE, .index = 6 }, .dim = 1 },
+	.quote_x = { .fg = { .type = PALETTE, .index = 6 }, .dim = 1 },
+};
+
+static inline int startswith(const char *s, const char *prefix)
+{
+	return !strncmp(s, prefix, strlen(prefix));
+}
+
+#define ARRAY_SIZE(a) (sizeof(a) / sizeof((a)[0]))
+
+static struct { const char *n; uint32_t c; } color_names[] = {
+	{"aliceblue", 0xf0f8ff}, {"antiquewhite", 0xfaebd7}, {"aqua", 0x00ffff},
+	{"aquamarine", 0x7fffd4}, {"azure", 0xf0ffff}, {"beige", 0xf5f5dc},
+	{"bisque", 0xffe4c4}, {"black", 0x000000}, {"blanchedalmond", 0xffebcd},
+	{"blue", 0x0000ff}, {"blueviolet", 0x8a2be2}, {"brown", 0xa52a2a},
+	{"burlywood", 0xdeb887}, {"cadetblue", 0x5f9ea0}, {"chartreuse", 0x7fff00},
+	{"chocolate", 0xd2691e}, {"coral", 0xff7f50}, {"cornflowerblue", 0x6495ed},
+	{"cornsilk", 0xfff8dc}, {"crimson", 0xdc143c}, {"darkblue", 0x00008b},
+	{"darkcyan", 0x008b8b}, {"darkgoldenrod", 0xb8860b}, {"darkgray", 0xa9a9a9},
+	{"darkgreen", 0x006400}, {"darkkhaki", 0xbdb76b}, {"darkmagenta", 0x8b008b},
+	{"darkolivegreen", 0x556b2f}, {"darkorange", 0xff8c00}, {"darkorchid", 0x9932cc},
+	{"darkred", 0x8b0000}, {"darksalmon", 0xe9967a}, {"darkseagreen", 0x8fbc8f},
+	{"darkslateblue", 0x483d8b}, {"darkslategray", 0x2f4f4f}, {"darkturquoise", 0x00ced1},
+	{"darkviolet", 0x9400d3}, {"deeppink", 0xff1493}, {"deepskyblue", 0x00bfff},
+	{"dimgray", 0x696969}, {"dodgerblue", 0x1e90ff}, {"firebrick", 0xb22222},
+	{"floralwhite", 0xfffaf0}, {"forestgreen", 0x228b22}, {"fuchsia", 0xff00ff},
+	{"gainsboro", 0xdcdcdc}, {"ghostwhite", 0xf8f8ff}, {"gold", 0xffd700},
+	{"goldenrod", 0xdaa520}, {"gray", 0x808080}, {"green", 0x008000},
+	{"greenyellow", 0xadff2f}, {"honeydew", 0xf0fff0}, {"hotpink", 0xff69b4},
+	{"indianred", 0xcd5c5c}, {"indigo", 0x4b0082}, {"ivory", 0xfffff0},
+	{"khaki", 0xf0e68c}, {"lavender", 0xe6e6fa}, {"lavenderblush", 0xfff0f5},
+	{"lawngreen", 0x7cfc00}, {"lemonchiffon", 0xfffacd}, {"lightblue", 0xadd8e6},
+	{"lightcoral", 0xf08080}, {"lightcyan", 0xe0ffff}, {"lightgoldenrodyellow", 0xfafad2},
+	{"lightgray", 0xd3d3d3}, {"lightgreen", 0x90ee90}, {"lightpink", 0xffb6c1},
+	{"lightsalmon", 0xffa07a}, {"lightseagreen", 0x20b2aa}, {"lightskyblue", 0x87cefa},
+	{"lightslategray", 0x778899}, {"lightsteelblue", 0xb0c4de}, {"lightyellow", 0xffffe0},
+	{"lime", 0x00ff00}, {"limegreen", 0x32cd32}, {"linen", 0xfaf0e6},
+	{"maroon", 0x800000}, {"mediumaquamarine", 0x66cdaa}, {"mediumblue", 0x0000cd},
+	{"mediumorchid", 0xba55d3}, {"mediumpurple", 0x9370db}, {"mediumseagreen", 0x3cb371},
+	{"mediumslateblue", 0x7b68ee}, {"mediumspringgreen", 0x00fa9a}, {"mediumturquoise", 0x48d1cc},
+	{"mediumvioletred", 0xc71585}, {"midnightblue", 0x191970}, {"mintcream", 0xf5fffa},
+	{"mistyrose", 0xffe4e1}, {"moccasin", 0xffe4b5}, {"navajowhite", 0xffdead},
+	{"navy", 0x000080}, {"oldlace", 0xfdf5e6}, {"olive", 0x808000},
+	{"olivedrab", 0x6b8e23}, {"orange", 0xffa500}, {"orangered", 0xff4500},
+	{"orchid", 0xda70d6}, {"palegoldenrod", 0xeee8aa}, {"palegreen", 0x98fb98},
+	{"paleturquoise", 0xafeeee}, {"palevioletred", 0xdb7093}, {"papayawhip", 0xffefd5},
+	{"peachpuff", 0xffdab9}, {"peru", 0xcd853f}, {"pink", 0xffc0cb},
+	{"plum", 0xdda0dd}, {"powderblue", 0xb0e0e6}, {"purple", 0x800080},
+	{"rebeccapurple", 0x663399}, {"red", 0xff0000}, {"rosybrown", 0xbc8f8f},
+	{"royalblue", 0x4169e1}, {"saddlebrown", 0x8b4513}, {"salmon", 0xfa8072},
+	{"sandybrown", 0xf4a460}, {"seagreen", 0x2e8b57}, {"seashell", 0xfff5ee},
+	{"sienna", 0xa0522d}, {"silver", 0xc0c0c0}, {"skyblue", 0x87ceeb},
+	{"slateblue", 0x6a5acd}, {"slategray", 0x708090}, {"snow", 0xfffafa},
+	{"springgreen", 0x00ff7f}, {"steelblue", 0x4682b4}, {"tan", 0xd2b48c},
+	{"teal", 0x008080}, {"thistle", 0xd8bfd8}, {"tomato", 0xff6347},
+	{"turquoise", 0x40e0d0}, {"violet", 0xee82ee}, {"wheat", 0xf5deb3},
+	{"white", 0xffffff}, {"whitesmoke", 0xf5f5f5}, {"yellow", 0xffff00},
+	{"yellowgreen", 0x9acd32},
+};
+
+static int color_name(const char *name, uint32_t *color)
+{
+	for (size_t c = 0; c < ARRAY_SIZE(color_names); c++) {
+		if (!strcmp(name, color_names[c].n)) {
+			*color = color_names[c].c;
+			return 0;
+		}
+	}
+	return 1;
+}
+
+static int parse_color(struct color *c, const char *val)
+{
+	uint32_t color = 0;
+	if (!strcmp(val, "default")) {
+		c->type = DEFAULT;
+	} else if (sscanf(val, "#%x", &color) == 1 && color <= 0xffffff) {
+		c->type = RGB;
+		c->rgb = color;
+	} else if (sscanf(val, "%d", &color) == 1 && color <= 256) {
+		c->type = PALETTE;
+		c->index = color;
+	} else if (!color_name(val, &color)) {
+		c->type = RGB;
+		c->rgb = color;
+	} else {
+		fprintf(stderr, "error: invalid color value %s\n", val);
+		return 1;
+	}
+	return 0;
+}
+
+static int parse_bool(int *b, const char *val)
+{
+	if (!strcmp(val, "true")) {
+		*b = 1;
+	} else if (!strcmp(val, "false")) {
+		*b = 0;
+	} else if (!strcmp(val, "toggle")) {
+		*b = !*b;
+	} else {
+		fprintf(stderr, "error: invalid bool value %s\n", val);
+		return 1;
+	}
+	return 0;
+}
+
+static int set_attr(struct style *s, const char *attr, const char *val)
+{
+	if (!strcmp(attr, "fg")) {
+		if (parse_color(&s->fg, val))
+			return 1;
+	} else if (!strcmp(attr, "bg")) {
+		if (parse_color(&s->fg, val))
+			return 1;
+	} else if (!strcmp(attr, "bold")) {
+		if (parse_bool(&s->bold, val))
+			return 1;
+	} else if (!strcmp(attr, "blink")) {
+		if (parse_bool(&s->blink, val))
+			return 1;
+	} else if (!strcmp(attr, "underline")) {
+		if (parse_bool(&s->underline, val))
+			return 1;
+	} else if (!strcmp(attr, "reverse")) {
+		if (parse_bool(&s->reverse, val))
+			return 1;
+	} else if (!strcmp(attr, "italic")) {
+		if (parse_bool(&s->italic, val))
+			return 1;
+	} else if (!strcmp(attr, "dim")) {
+		if (parse_bool(&s->dim, val))
+			return 1;
+	} else if (!strcmp(attr, "normal")) {
+		s->bold = 0;
+		s->underline = 0;
+		s->reverse = 0;
+		s->italic = 0;
+		s->dim = 0;
+	} else if (!strcmp(attr, "default")) {
+		s->fg.type = NONE;
+		s->fg.type = NONE;
+	} else {
+		fprintf(stderr, "error: invalid style attribute %s\n", attr);
+		return 1;
+	}
+	return 0;
+}
+
+static struct {const char *n; struct style *s;} ini_objects[] = {
+	{"url", &styles.url},
+	{"header", &styles.header},
+	{"signature", &styles.signature},
+	{"diff_meta", &styles.diff_meta},
+	{"diff_chunk", &styles.diff_chunk},
+	{"diff_add", &styles.diff_add},
+	{"diff_del", &styles.diff_del},
+	{"quote_1", &styles.quote_1},
+	{"quote_2", &styles.quote_2},
+	{"quote_3", &styles.quote_3},
+	{"quote_4", &styles.quote_4},
+	{"quote_x", &styles.quote_x},
+};
+
+static int parse_styleset(void)
+{
+	int in_section = 0;
+	char buf[BUFSIZ];
+	int err = 0;
+	FILE *f;
+
+	if (!styleset)
+		return 0;
+
+	f = fopen(styleset, "r");
+	if (!f) {
+		perror("error: failed to open styleset");
+		return 1;
+	}
+
+	while (fgets(buf, sizeof(buf), f)) {
+		/* strip LF, CR, CRLF, LFCR */
+		buf[strcspn(buf, "\r\n")] = '\0';
+		if (in_section) {
+			char obj[64], attr[64], val[64];
+			int changed = 0;
+
+			if (sscanf(buf, "%63[^.].%63[^=] = %63s", obj, attr, val) != 3)
+				continue;
+
+			for (size_t o = 0; o < ARRAY_SIZE(ini_objects); o++) {
+				if (fnmatch(obj, ini_objects[o].n, 0))
+					continue;
+				if (set_attr(ini_objects[o].s, attr, val)) {
+					err = 1;
+					goto end;
+				}
+				changed++;
+			}
+			if (!changed) {
+				fprintf(stderr,
+					"error: unknown style object %s\n",
+					obj);
+				err = 1;
+				goto end;
+			}
+		} else if (!strcmp(buf, "[viewer]")) {
+			in_section = 1;
+			/* only disable the default theme if there is
+			 * a [viewer] section in the styleset */
+			memset(&styles, 0, sizeof(styles));
+		}
+	}
+
+end:
+	fclose(f);
+	return err;
+}
+
+static inline void print(const char *in)
+{
+	fputs(in, stdout);
+}
+
+static inline size_t print_notabs(const char *in, size_t max_len)
+{
+	size_t len = 0;
+	while (*in != '\0' && len < max_len) {
+		char c = *in++;
+		if (c == '\t') {
+			/* Tabs are interpreted as cursor movement and are not
+			 * colored like regular characters. Replace them with
+			 * 8 spaces. */
+			fputs("        ", stdout);
+		} else {
+			fputc(c, stdout);
+		}
+		len++;
+	}
+	return len;
+}
+
+static void diff_chunk(const char *in)
+{
+	size_t len = 0;
+	print(seq(&styles.diff_chunk));
+	while (in[len] == '@')
+		len++;
+	while (in[len] != '\0' && in[len] != '@')
+		len++;
+	while (in[len] == '@')
+		len++;
+	in += print_notabs(in, len);
+	print(RESET);
+	print_notabs(in, BUFSIZ);
+}
+
+#define URL_RE \
+	"[a-z]{2,8}://[[:graph:]]{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];
+	size_t len;
+	int trim;
+
+	while (!regexec(&url_re, in, 2, groups, 0)) {
+		in += print_notabs(in, groups[0].rm_so);
+		print(seq(&styles.url));
+		len = groups[0].rm_eo - groups[0].rm_so;
+		/* Heuristic to remove trailing characters that are valid URL
+		 * characters, but typically not at the end of the URL */
+		trim = 1;
+		while (trim && len > 0) {
+			switch (in[len - 1]) {
+			case '>': case '.': case ',': case ';': case ')':
+			case '!': case '?': case '"': case '\'':
+				len--;
+				break;
+			default:
+				trim = 0;
+				break;
+			}
+		}
+		in += print_notabs(in, len);
+		print(RESET);
+		if (ctx) {
+			print(seq(ctx));
+		}
+	}
+	print_notabs(in, BUFSIZ);
+}
+
+static inline void signature(const char *in)
+{
+	print(seq(&styles.signature));
+	urls(in, &styles.signature);
+	print(RESET);
+}
+
+#define HEADER_RE "^[A-Z][[:alnum:]_-]+:"
+static regex_t header_re;
+
+static void header(const char *in)
+{
+	regmatch_t groups[1];
+
+	if (!regexec(&header_re, in, 1, groups, 0)) {
+		print(seq(&styles.header));
+		in += print_notabs(in, groups[0].rm_eo);
+		print(RESET);
+	}
+	urls(in, NULL);
+}
+
+#define DIFF_META_RE \
+	"^(diff --git|(new|deleted) file|similarity" \
+	" index|(rename|copy) (to|from)|index|---|\\+\\+\\+) "
+static regex_t diff_meta_re;
+
+static void quote(const char *in)
+{
+	regmatch_t groups[8];
+	struct style *s;
+	int q, level;
+
+	q = level = 0;
+	while (in[q] == '>') {
+		level++;
+		q++;
+		if (in[q] == ' ')
+			q++;
+	}
+	switch (level) {
+	case 1:
+		s = &styles.quote_1;
+		break;
+	case 2:
+		s = &styles.quote_2;
+		break;
+	case 3:
+		s = &styles.quote_3;
+		break;
+	case 4:
+		s = &styles.quote_4;
+		break;
+	default:
+		s = &styles.quote_x;
+		break;
+	}
+
+	print(seq(s));
+	in += print_notabs(in, q);
+	if (startswith(in, "+")) {
+		printf("%s%s", RESET, seq(&styles.diff_add));
+		print_notabs(in, BUFSIZ);
+	} else if (startswith(in, "-")) {
+		printf("%s%s", RESET, seq(&styles.diff_del));
+		print_notabs(in, BUFSIZ);
+	} else if (!regexec(&diff_meta_re, in, 8, groups, 0)) {
+		print(BOLD);
+		print_notabs(in, BUFSIZ);
+	} else {
+		urls(in, s);
+	}
+	print(RESET);
+}
+
+static void print_style(const char *in, struct style *s)
+{
+	print(seq(s));
+	print_notabs(in, BUFSIZ);
+	print(RESET);
+}
+
+enum state { INIT, DIFF, SIGNATURE, BODY };
+
+static void colorize_line(const char *in)
+{
+	static enum state state = INIT;
+	regmatch_t groups[8];  /* enough groups to cover all expressions */
+
+	switch (state) {
+	case DIFF:
+		if (!strcmp(in, "--") || !strcmp(in, "-- ")) {
+			state = SIGNATURE;
+			signature(in);
+		} else if (startswith(in, "@@ ")) {
+			diff_chunk(in);
+		} else if (!regexec(&diff_meta_re, in, 8, groups, 0)) {
+			print_style(in, &styles.diff_meta);
+		} else if (startswith(in, "+")) {
+			print_style(in, &styles.diff_add);
+		} else if (startswith(in, "-")) {
+			print_style(in, &styles.diff_del);
+		} else if (!startswith(in, " ") && strcmp(in, "") != 0) {
+			state = BODY;
+			if (startswith(in, ">")) {
+				quote(in);
+			} else {
+				urls(in, NULL);
+			}
+		} else {
+			print_notabs(in, BUFSIZ);
+		}
+		break;
+	case SIGNATURE:
+		signature(in);
+		break;
+	case BODY:
+		if (startswith(in, ">")) {
+			quote(in);
+		} else if (startswith(in, "diff --git ")) {
+			state = DIFF;
+			print_style(in, &styles.diff_meta);
+		} else if (!strcmp(in, "--") || !strcmp(in, "-- ")) {
+			state = SIGNATURE;
+			signature(in);
+		} else if (!regexec(&header_re, in, 8, groups, 0)) {
+			header(in);
+		} else {
+			urls(in, NULL);
+		}
+		break;
+	default: /* INIT */
+		if (startswith(in, "diff --git ")) {
+			state = DIFF;
+			print_style(in, &styles.diff_meta);
+		} else if (!strcmp(in, "--") || !strcmp(in, "-- ")) {
+			state = SIGNATURE;
+			signature(in);
+		} else {
+			state = BODY;
+			if (startswith(in, ">")) {
+				quote(in);
+			} else if (!regexec(&header_re, in, 8, groups, 0)) {
+				header(in);
+			} else {
+				urls(in, NULL);
+			}
+		}
+		break;
+	}
+}
+
+int parse_args(int argc, char **argv)
+{
+	const char *filename = NULL;
+	char c;
+
+	styleset = getenv("AERC_STYLESET");
+
+	while ((c = getopt(argc, argv, "hs:f:")) != -1) {
+		switch (c) {
+		case 's':
+			styleset = optarg;
+			break;
+		case 'f':
+			filename = optarg;
+			break;
+		default:
+			usage();
+			return 1;
+		}
+	}
+	if (optind < argc) {
+		fprintf(stderr, "%s: unexpected argument -- '%s'\n",
+			argv[0], argv[optind]);
+		usage();
+		return 1;
+	}
+	if (filename == NULL || !strcmp(filename, "-")) {
+		in_file = stdin;
+	} else {
+		in_file = fopen(filename, "r");
+		if (!in_file) {
+			perror("error: cannot open file");
+			return 1;
+		}
+	}
+	return 0;
+}
+
+int main(int argc, char **argv)
+{
+	char buf[BUFSIZ];
+	int err;
+
+	regcomp(&header_re, HEADER_RE, REG_EXTENDED);
+	regcomp(&diff_meta_re, DIFF_META_RE, REG_EXTENDED);
+	regcomp(&url_re, URL_RE, REG_EXTENDED);
+
+	err = parse_args(argc, argv);
+	if (err) {
+		goto end;
+	}
+	err = parse_styleset();
+	if (err) {
+		goto end;
+	}
+	while (fgets(buf, sizeof(buf), in_file)) {
+		/* strip LF, CR, CRLF, LFCR */
+		buf[strcspn(buf, "\r\n")] = '\0';
+		colorize_line(buf);
+		printf("\n");
+	}
+end:
+	if (in_file) {
+		fclose(in_file);
+	}
+	return err;
+}
diff --git a/filters/vectors/colorize-patch.expected b/filters/vectors/colorize-patch.expected
new file mode 100644
index 00000000..1a2144a0
--- /dev/null
+++ b/filters/vectors/colorize-patch.expected
@@ -0,0 +1,49 @@
+From: Robin Jarry <robin@jarry.cc>
+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>
+---
+ doc/aerc-stylesets.7.scd | 18 +++++++++---------
+ 1 file changed, 9 insertions(+), 9 deletions(-)
+
+diff --git a/doc/aerc-stylesets.7.scd b/doc/aerc-stylesets.7.scd
+index d82ba7cf8163..34bbf4af0fc5 100644
+--- a/doc/aerc-stylesets.7.scd
++++ b/doc/aerc-stylesets.7.scd
+@@ -180,20 +180,20 @@ that style applies, unless overridden by a higher layer.
+ 
+ The order that *msglist_\** styles are applied in is, from first to last:
+ 
+-1. *msglist_default*
+-2. *msglist_unread*
+-3. *msglist_read*
+-4. *msglist_flagged*
+-5. *msglist_deleted*
+-6. *msglist_marked*
++. *msglist_default*
++. *msglist_unread*
++. *msglist_read*
++. *msglist_flagged*
++. *msglist_deleted*
++. *msglist_marked*
+ 
+ So, the marked style will override all other msglist styles.
+ 
+ The order for *dirlist_\** styles is:
+ 
+-1. *dirlist_default*
+-2. *dirlist_unread*
+-3. *dirlist_recent*
++. *dirlist_default*
++. *dirlist_unread*
++. *dirlist_recent*
+ 
+ ## COLORS
+ 
+-- 
+2.39.0
+
diff --git a/filters/vectors/colorize-patch.in b/filters/vectors/colorize-patch.in
new file mode 100644
index 00000000..48e12d8e
--- /dev/null
+++ b/filters/vectors/colorize-patch.in
@@ -0,0 +1,49 @@
+From: Robin Jarry <robin@jarry.cc>
+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>
+---
+ doc/aerc-stylesets.7.scd | 18 +++++++++---------
+ 1 file changed, 9 insertions(+), 9 deletions(-)
+
+diff --git a/doc/aerc-stylesets.7.scd b/doc/aerc-stylesets.7.scd
+index d82ba7cf8163..34bbf4af0fc5 100644
+--- a/doc/aerc-stylesets.7.scd
++++ b/doc/aerc-stylesets.7.scd
+@@ -180,20 +180,20 @@ that style applies, unless overridden by a higher layer.
+ 
+ The order that *msglist_\** styles are applied in is, from first to last:
+ 
+-1. *msglist_default*
+-2. *msglist_unread*
+-3. *msglist_read*
+-4. *msglist_flagged*
+-5. *msglist_deleted*
+-6. *msglist_marked*
++. *msglist_default*
++. *msglist_unread*
++. *msglist_read*
++. *msglist_flagged*
++. *msglist_deleted*
++. *msglist_marked*
+ 
+ So, the marked style will override all other msglist styles.
+ 
+ The order for *dirlist_\** styles is:
+ 
+-1. *dirlist_default*
+-2. *dirlist_unread*
+-3. *dirlist_recent*
++. *dirlist_default*
++. *dirlist_unread*
++. *dirlist_recent*
+ 
+ ## COLORS
+ 
+-- 
+2.39.0
+
diff --git a/filters/vectors/colorize-quotes.expected b/filters/vectors/colorize-quotes.expected
new file mode 100644
index 00000000..d350e4f1
--- /dev/null
+++ b/filters/vectors/colorize-quotes.expected
@@ -0,0 +1,65 @@
+Foo Bar, xxxxx:
+> Lorem ipsum dolor sit amet, insolens adolescens ne usu? In pri denique
+> argumentum, te autem decore convenire mea! Duo nisl esse an, aliquid
+> conceptam sea cu. Ignota copiosae gubergren ad est, ut illum doming vocibus
+> sed. Et vis nulla expetendis mediocritatem, errem option gloriatur at nam?
+> Brute vidisse corpora ut his, sonet omnesque adipiscing ea quo, cum ea errem
+> aliquip reformidans?
+
+Magna delicatissimi ei vel? Quem petentium scribentur eum ne? Et inani debet
+cetero mea, sint conceptam efficiendi mel te. Qui ut senserit interesset, per
+nibh petentium at! Sit docendi laboramus ei, animal insolens ad mea.
+
+>> Nostrud alienum nec in, illum errem audiam no per! Saepe alterum vis ea! Ei
+>> quis minim ius, ut eos mandamus salutandi. Lorem facilisis in nam, ridens
+>> principes sadipscing et eum, pri graecis singulis ut. Mea dolor primis
+>> impetus in, his epicurei tacimates id, vis labitur suscipit ad.
+> 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/".
+>
+> Error libris deleniti ea mei, vis at elit probo munere, his sint unum
+> albucius ex.
+
+Graece definiebas scripserit ne est? Nec nonumes explicari contentiones ne,
+vocent iuvaret placerat no vix. Nec et partem salutandi deseruisse, his no
+possim malorum pericula. Te quando reprehendunt nam, at consul sadipscing vel?
+Velit possim aliquando ei per, ne simul quodsi antiopam sea, ullum choro
+facilisi et pri!
+
+> Dico soleat partem ea pro, ad vix impetus splendide. Primis melius principes
+> pri ad, tacimates pertinacia ei pro? Appareat atomorum oportere at nam, eu
+> per quod minim reprimique, ornatus graecis ad vel. Malis vulputate ea qui,
+> eum tacimates recteque et, usu ea dolore vidisse. Brute mediocrem molestiae
+> sed te. No stet prompta pri, rebum populo nominati eos te.
+>
+> diff --git a/foo b/foo
+> index 4b0fe8dded3a..518b67134639 100644
+> --- a/foo
+> +++ b/foo
+> @@ -131,6 +131,83 @@ func pouet() int {
+>          err := doThis()
+>  
+> -        err2 := doThat()
+> +        err2 := notDoThat()
+>  
+>          if err != nil || err2 != nil {
+
+Id vix referrentur philosophia, veri labores an nec. Noster denique no duo, sit
+ei diam inermis vocibus! Mutat principes ex pro, at pericula assueverit vel.
+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==
+
+Eam mundi libris debitis ad, eam regione numquam at. Eum omnes bonorum eu,
+oporteat assueverit disputationi nam ne, nonumes iracundia mea ad! Duo libris
+recusabo id, ceteros salutatus inciderint vim ea. Et graeco reformidans vel? Ei
+has labore quidam?
+
+>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> sympa, non?
+
+-- 
+Batman
diff --git a/filters/vectors/colorize-quotes.in b/filters/vectors/colorize-quotes.in
new file mode 100644
index 00000000..31b04911
--- /dev/null
+++ b/filters/vectors/colorize-quotes.in
@@ -0,0 +1,65 @@
+Foo Bar, xxxxx:
+> Lorem ipsum dolor sit amet, insolens adolescens ne usu? In pri denique
+> argumentum, te autem decore convenire mea! Duo nisl esse an, aliquid
+> conceptam sea cu. Ignota copiosae gubergren ad est, ut illum doming vocibus
+> sed. Et vis nulla expetendis mediocritatem, errem option gloriatur at nam?
+> Brute vidisse corpora ut his, sonet omnesque adipiscing ea quo, cum ea errem
+> aliquip reformidans?
+
+Magna delicatissimi ei vel? Quem petentium scribentur eum ne? Et inani debet
+cetero mea, sint conceptam efficiendi mel te. Qui ut senserit interesset, per
+nibh petentium at! Sit docendi laboramus ei, animal insolens ad mea.
+
+>> Nostrud alienum nec in, illum errem audiam no per! Saepe alterum vis ea! Ei
+>> quis minim ius, ut eos mandamus salutandi. Lorem facilisis in nam, ridens
+>> principes sadipscing et eum, pri graecis singulis ut. Mea dolor primis
+>> impetus in, his epicurei tacimates id, vis labitur suscipit ad.
+> 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/".
+>
+> Error libris deleniti ea mei, vis at elit probo munere, his sint unum
+> albucius ex.
+
+Graece definiebas scripserit ne est? Nec nonumes explicari contentiones ne,
+vocent iuvaret placerat no vix. Nec et partem salutandi deseruisse, his no
+possim malorum pericula. Te quando reprehendunt nam, at consul sadipscing vel?
+Velit possim aliquando ei per, ne simul quodsi antiopam sea, ullum choro
+facilisi et pri!
+
+> Dico soleat partem ea pro, ad vix impetus splendide. Primis melius principes
+> pri ad, tacimates pertinacia ei pro? Appareat atomorum oportere at nam, eu
+> per quod minim reprimique, ornatus graecis ad vel. Malis vulputate ea qui,
+> eum tacimates recteque et, usu ea dolore vidisse. Brute mediocrem molestiae
+> sed te. No stet prompta pri, rebum populo nominati eos te.
+>
+> diff --git a/foo b/foo
+> index 4b0fe8dded3a..518b67134639 100644
+> --- a/foo
+> +++ b/foo
+> @@ -131,6 +131,83 @@ func pouet() int {
+>  	err := doThis()
+>  
+> -	err2 := doThat()
+> +	err2 := notDoThat()
+>  
+>  	if err != nil || err2 != nil {
+
+Id vix referrentur philosophia, veri labores an nec. Noster denique no duo, sit
+ei diam inermis vocibus! Mutat principes ex pro, at pericula assueverit vel.
+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==
+
+Eam mundi libris debitis ad, eam regione numquam at. Eum omnes bonorum eu,
+oporteat assueverit disputationi nam ne, nonumes iracundia mea ad! Duo libris
+recusabo id, ceteros salutatus inciderint vim ea. Et graeco reformidans vel? Ei
+has labore quidam?
+
+>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> sympa, non?
+
+-- 
+Batman