From 577381b3c90a2b7593ed80cb85be1e34016faf2f Mon Sep 17 00:00:00 2001
From: Catherine <whitequark@whitequark.org>
Date: Thu, 28 Mar 2024 10:06:18 +0000
Subject: [PATCH] fmt: if enabled, group padding zeroes.

Before this commit, the combination of `_` and `0` format characters
would produce a result like `000000001010_1010`.
After this commit, it would be `0000_0000_1010_1010`.

This has a slight quirk where a format like `{:020_b}` results in
the output `0_0000_0000_1010_1010`, which is one character longer than
requested. Python has the same behavior, and it's not clear what would
be strictly speaking correct, so Python behavior is implemented.
---
 backends/cxxrtl/runtime/cxxrtl/cxxrtl.h | 22 ++++++++++++++--------
 kernel/fmt.cc                           | 12 +++++++++---
 2 files changed, 23 insertions(+), 11 deletions(-)

diff --git a/backends/cxxrtl/runtime/cxxrtl/cxxrtl.h b/backends/cxxrtl/runtime/cxxrtl/cxxrtl.h
index f628051105e..bc26f2b2e03 100644
--- a/backends/cxxrtl/runtime/cxxrtl/cxxrtl.h
+++ b/backends/cxxrtl/runtime/cxxrtl/cxxrtl.h
@@ -1109,28 +1109,27 @@ struct fmt_part {
 					}
 				}
 
-				size_t width = Bits;
+				size_t val_width = Bits;
 				if (base != 10) {
-					width = 1;
+					val_width = 1;
 					for (size_t index = 0; index < Bits; index++)
 						if (val.bit(index))
-							width = index + 1;
+							val_width = index + 1;
 				}
 
 				if (base == 2) {
 					if (show_base)
 						prefix += "0b";
-					for (size_t index = 0; index < width; index++) {
+					for (size_t index = 0; index < val_width; index++) {
 						if (group && index > 0 && index % 4 == 0)
 							buf += '_';
 						buf += (val.bit(index) ? '1' : '0');
 					}
-					std::reverse(buf.begin(), buf.end());
 				} else if (base == 8 || base == 16) {
 					if (show_base)
 						prefix += (base == 16) ? (hex_upper ? "0X" : "0x") : "0o";
 					size_t step = (base == 16) ? 4 : 3;
-					for (size_t index = 0; index < width; index += step) {
+					for (size_t index = 0; index < val_width; index += step) {
 						if (group && index > 0 && index % (4 * step) == 0)
 							buf += '_';
 						uint8_t value = val.bit(index) | (val.bit(index + 1) << 1) | (val.bit(index + 2) << 2);
@@ -1138,7 +1137,6 @@ struct fmt_part {
 							value |= val.bit(index + 3) << 3;
 						buf += (hex_upper ? "0123456789ABCDEF" : "0123456789abcdef")[value];
 					}
-					std::reverse(buf.begin(), buf.end());
 				} else if (base == 10) {
 					if (show_base)
 						prefix += "0d";
@@ -1158,8 +1156,16 @@ struct fmt_part {
 						xval = quotient;
 						index++;
 					}
-					std::reverse(buf.begin(), buf.end());
 				} else assert(false && "Unsupported base for fmt_part");
+				if (justify == NUMERIC && group && padding == '0') {
+					int group_size = base == 10 ? 3 : 4;
+					while (prefix.size() + buf.size() < width) {
+						if (buf.size() % (group_size + 1) == group_size)
+							buf += '_';
+						buf += '0';
+					}
+				}
+				std::reverse(buf.begin(), buf.end());
 				break;
 			}
 
diff --git a/kernel/fmt.cc b/kernel/fmt.cc
index 43d8feddfa4..cbf2d12e9ac 100644
--- a/kernel/fmt.cc
+++ b/kernel/fmt.cc
@@ -776,7 +776,6 @@ std::string Fmt::render() const
 							else /* if (bit == State::S0) */
 								buf += '0';
 						}
-						std::reverse(buf.begin(), buf.end());
 					} else if (part.base == 8 || part.base == 16) {
 						if (part.show_base)
 							prefix += (part.base == 16) ? (part.hex_upper ? "0X" : "0x") : "0o";
@@ -807,7 +806,6 @@ std::string Fmt::render() const
 							else
 								buf += (part.hex_upper ? "0123456789ABCDEF" : "0123456789abcdef")[subvalue.as_int()];
 						}
-						std::reverse(buf.begin(), buf.end());
 					} else if (part.base == 10) {
 						if (part.show_base)
 							prefix += "0d";
@@ -831,9 +829,17 @@ std::string Fmt::render() const
 								value = RTLIL::const_div(value, 10, false, false, value.size());
 								index++;
 							}
-							std::reverse(buf.begin(), buf.end());
 						}
 					} else log_abort();
+					if (part.justify == FmtPart::NUMERIC && part.group && part.padding == '0') {
+						int group_size = part.base == 10 ? 3 : 4;
+						while (prefix.size() + buf.size() < part.width) {
+							if (buf.size() % (group_size + 1) == group_size)
+								buf += '_';
+							buf += '0';
+						}
+					}
+					std::reverse(buf.begin(), buf.end());
 				} else if (part.type == FmtPart::STRING) {
 					buf = part.sig.as_const().decode_string();
 				} else if (part.type == FmtPart::VLOG_TIME) {