From fff5211d2776185348d6b813f2a16cedd5f485d0 Mon Sep 17 00:00:00 2001
From: Phil Jones <philj56@gmail.com>
Date: Thu, 1 Dec 2022 20:24:35 +0000
Subject: Add --ascii-input option.

---
 README.md        |  5 +++++
 completions/tofi |  1 +
 doc/config       |  5 +++++
 doc/tofi.5.md    |  8 ++++++++
 doc/tofi.5.scd   |  7 +++++++
 src/config.c     |  5 +++++
 src/main.c       | 16 +++++++++++++---
 src/tofi.h       |  1 +
 8 files changed, 45 insertions(+), 3 deletions(-)

diff --git a/README.md b/README.md
index 6344599..fee5463 100644
--- a/README.md
+++ b/README.md
@@ -181,6 +181,11 @@ In roughly descending order, the most important options for performance are:
   sharpness, but your mileage may vary. This option has no effect if a path to
   a font file hasn't been passed to `--font`.
 
+* `--ascii-input` - Proper Unicode handling is slower than plain ASCII - on the
+  order of a few ms for ~40 kB of input. Specifying `--ascii-input true` will
+  disable some of this handling, speeding up tofi's startup, but searching for
+  non-ASCII characters may not work properly.
+
 * `--late-keyboard-init` - The last avoidable thing that slows down startup is
   initialisation of the keyboard. This only takes 1-2ms on my laptop, but up
   to 60ms on a Raspberry Pi Zero 2 W. Passing this option will delay keyboard
diff --git a/completions/tofi b/completions/tofi
index 328f7dc..d871db1 100644
--- a/completions/tofi
+++ b/completions/tofi
@@ -79,6 +79,7 @@ _tofi()
 		--hint-font
 		--late-keyboard-init
 		--multi-instance
+		--ascii-input
      )
 
 	case "${prev}" in
diff --git a/doc/config b/doc/config
index 017aeb1..4ce0a7d 100644
--- a/doc/config
+++ b/doc/config
@@ -255,6 +255,11 @@
 	# from running simultaneously.
 	multi-instance = false
 
+	# Assume input is plain ASCII, and disable some Unicode handling
+	# functions. This is faster, but means e.g. a search for "e" will not
+	# match "é".
+	ascii-input = false
+
 #
 ### Inclusion
 #
diff --git a/doc/tofi.5.md b/doc/tofi.5.md
index e8f3c49..622858f 100644
--- a/doc/tofi.5.md
+++ b/doc/tofi.5.md
@@ -137,6 +137,14 @@ options.
 >
 > Default: false
 
+**ascii-input**=*true\|false*
+
+> Assume input is plain ASCII, and disable some Unicode handling
+> functions. This is faster, but means e.g. a search for "e" will not
+> match "é".
+>
+> Default: false
+
 # STYLE OPTIONS
 
 **font**=*font*
diff --git a/doc/tofi.5.scd b/doc/tofi.5.scd
index 72be59e..4dde4d6 100644
--- a/doc/tofi.5.scd
+++ b/doc/tofi.5.scd
@@ -118,6 +118,13 @@ options.
 
 	Default: false
 
+*ascii-input*=_true|false_
+	Assume input is plain ASCII, and disable some Unicode handling
+	functions. This is faster, but means e.g. a search for "e" will not
+	match "é".
+
+	Default: false
+
 # STYLE OPTIONS
 
 *font*=_font_
diff --git a/src/config.c b/src/config.c
index 83321ed..0cfd566 100644
--- a/src/config.c
+++ b/src/config.c
@@ -686,6 +686,11 @@ bool parse_option(struct tofi *tofi, const char *filename, size_t lineno, const
 		if (!err) {
 			tofi->multiple_instance = val;
 		}
+	} else if (strcasecmp(option, "ascii-input") == 0) {
+		bool val = parse_bool(filename, lineno, value, &err);
+		if (!err) {
+			tofi->ascii_input = val;
+		}
 	} else if (strcasecmp(option, "late-keyboard-init") == 0) {
 		bool val = parse_bool(filename, lineno, value, &err);
 		if (!err) {
diff --git a/src/main.c b/src/main.c
index 7a92c19..cde56b4 100644
--- a/src/main.c
+++ b/src/main.c
@@ -48,7 +48,7 @@ static uint32_t gettime_ms() {
 
 
 /* Read all of stdin into a buffer. */
-static char *read_stdin() {
+static char *read_stdin(bool normalize) {
 	const size_t block_size = BUFSIZ;
 	size_t num_blocks = 1;
 	size_t buf_size = block_size;
@@ -66,12 +66,21 @@ static char *read_stdin() {
 				stdin);
 		if (bytes_read != block_size) {
 			if (!feof(stdin) && ferror(stdin)) {
-				log_error("Error reading stdin\n");
+				log_error("Error reading stdin.\n");
 			}
 			buf[block * block_size + bytes_read] = '\0';
 			break;
 		}
 	}
+	if (normalize) {
+		if (utf8_validate(buf)) {
+			char *tmp = utf8_normalize(buf);
+			free(buf);
+			buf = tmp;
+		} else {
+			log_error("Invalid UTF-8 in stdin.\n");
+		}
+	}
 	return buf;
 }
 
@@ -875,6 +884,7 @@ const struct option long_options[] = {
 	{"terminal", required_argument, NULL, 0},
 	{"hint-font", required_argument, NULL, 0},
 	{"multi-instance", required_argument, NULL, 0},
+	{"ascii-input", required_argument, NULL, 0},
 	{"output", required_argument, NULL, 0},
 	{"scale", required_argument, NULL, 0},
 	{"late-keyboard-init", optional_argument, NULL, 'k'},
@@ -1360,7 +1370,7 @@ int main(int argc, char *argv[])
 		log_debug("App list generated.\n");
 	} else {
 		log_debug("Reading stdin.\n");
-		char *buf = read_stdin();
+		char *buf = read_stdin(!tofi.ascii_input);
 		tofi.window.entry.command_buffer = buf;
 		tofi.window.entry.commands = string_ref_vec_from_buffer(buf);
 		if (tofi.use_history) {
diff --git a/src/tofi.h b/src/tofi.h
index b7437a1..547b4ed 100644
--- a/src/tofi.h
+++ b/src/tofi.h
@@ -86,6 +86,7 @@ struct tofi {
 
 	/* Options */
 	uint32_t anchor;
+	bool ascii_input;
 	bool hide_cursor;
 	bool use_history;
 	bool use_scale;
-- 
cgit v1.2.3