From bc01468c7545b2362a52b48f11fa0e5cccb4e75a Mon Sep 17 00:00:00 2001
From: Jannis Harder <me@jix.one>
Date: Tue, 1 Apr 2025 13:50:29 +0200
Subject: [PATCH] read_liberty: Faster std::string construction in the liberty
 lexer

This extends the `LibertyInputStream` added in the previous commit to
allow arbitrary lookahead. Then this uses the lookahead to find the
total length of the token within the input buffer, instead of consuming
the token byte by byte while appending to a std::string. Constructing
the std::string with the total length is known avoids any reallocations
from growing std::string's buffer.
---
 passes/techmap/libparse.cc | 55 ++++++++++++++++++++++++--------------
 passes/techmap/libparse.h  | 11 ++++++++
 2 files changed, 46 insertions(+), 20 deletions(-)

diff --git a/passes/techmap/libparse.cc b/passes/techmap/libparse.cc
index dbf191080..d3d5b7d57 100644
--- a/passes/techmap/libparse.cc
+++ b/passes/techmap/libparse.cc
@@ -77,6 +77,16 @@ int LibertyInputStream::get_cold()
 	return c;
 }
 
+int LibertyInputStream::peek_cold(size_t offset)
+{
+	if (buf_pos + offset >= buf_end) {
+		if (!extend_buffer_at_least(offset + 1))
+			return EOF;
+	}
+
+	return buffer[buf_pos + offset];
+}
+
 LibertyAst::~LibertyAst()
 {
 	for (auto child : children)
@@ -282,15 +292,19 @@ int LibertyParser::lexer(std::string &str)
 
 	// search for identifiers, numbers, plus or minus.
 	if (('a' <= c && c <= 'z') || ('A' <= c && c <= 'Z') || ('0' <= c && c <= '9') || c == '_' || c == '-' || c == '+' || c == '.') {
-		str = static_cast<char>(c);
-		while (1) {
-			c = f.get();
+		f.unget();
+		size_t i = 1;
+		while (true) {
+			c = f.peek(i);
 			if (('a' <= c && c <= 'z') || ('A' <= c && c <= 'Z') || ('0' <= c && c <= '9') || c == '_' || c == '-' || c == '+' || c == '.')
-				str += c;
+				i += 1;
 			else
 				break;
 		}
-		f.unget();
+		str.clear();
+		str.append(f.buffered_data(), f.buffered_data() + i);
+		f.consume(i);
+
 		if (str == "+" || str == "-") {
 			/* Single operator is not an identifier */
 			// fprintf(stderr, "LEX: char >>%s<<\n", str.c_str());
@@ -305,23 +319,24 @@ int LibertyParser::lexer(std::string &str)
 	// if it wasn't an identifer, number of array range,
 	// maybe it's a string?
 	if (c == '"') {
-		str = "";
-#ifdef FILTERLIB
-		str += c;
-#endif
-		while (1) {
-			c = f.get();
-			if (c == '\n')
-				line++;
-			if (c == '"') {
-#ifdef FILTERLIB
-				str += c;
-#endif
+		size_t i = 0;
+		while (true) {
+			c = f.peek(i);
+			line += (c == '\n');
+			if (c != '"')
+				i += 1;
+			else
 				break;
-			}
-			str += c;
 		}
-		// fprintf(stderr, "LEX: string >>%s<<\n", str.c_str());
+		str.clear();
+#ifdef FILTERLIB
+		f.unget();
+		str.append(f.buffered_data(), f.buffered_data() + i + 2);
+		f.consume(i + 2);
+#else
+		str.append(f.buffered_data(), f.buffered_data() + i);
+		f.consume(i + 1);
+#endif
 		return 'v';
 	}
 
diff --git a/passes/techmap/libparse.h b/passes/techmap/libparse.h
index eb73e296d..1fcaaebee 100644
--- a/passes/techmap/libparse.h
+++ b/passes/techmap/libparse.h
@@ -101,6 +101,7 @@ namespace Yosys
 		bool extend_buffer_at_least(size_t size = 1);
 
 		YS_COLD int get_cold();
+		YS_COLD int peek_cold(size_t offset);
 
 	public:
 		LibertyInputStream(std::istream &f) : f(f) {}
@@ -116,6 +117,16 @@ namespace Yosys
 			return c;
 		}
 
+		int peek(size_t offset = 0) {
+			if (buf_pos + offset >= buf_end)
+				return peek_cold(offset);
+			return buffer[buf_pos + offset];
+		}
+
+		void consume(size_t n = 1) {
+			buf_pos += n;
+		}
+
 		void unget() {
 			buf_pos -= 1;
 		}