10b57cec5SDimitry Andric //===- FormatVariadic.cpp - Format string parsing and analysis ----*-C++-*-===//
20b57cec5SDimitry Andric //
30b57cec5SDimitry Andric // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
40b57cec5SDimitry Andric // See https://llvm.org/LICENSE.txt for license information.
50b57cec5SDimitry Andric // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
60b57cec5SDimitry Andric //===----------------------------------------------------------------------===//
70b57cec5SDimitry Andric
80b57cec5SDimitry Andric #include "llvm/Support/FormatVariadic.h"
95ffd83dbSDimitry Andric #include <cassert>
10bdd1243dSDimitry Andric #include <optional>
110b57cec5SDimitry Andric
120b57cec5SDimitry Andric using namespace llvm;
130b57cec5SDimitry Andric
translateLocChar(char C)14bdd1243dSDimitry Andric static std::optional<AlignStyle> translateLocChar(char C) {
150b57cec5SDimitry Andric switch (C) {
160b57cec5SDimitry Andric case '-':
170b57cec5SDimitry Andric return AlignStyle::Left;
180b57cec5SDimitry Andric case '=':
190b57cec5SDimitry Andric return AlignStyle::Center;
200b57cec5SDimitry Andric case '+':
210b57cec5SDimitry Andric return AlignStyle::Right;
220b57cec5SDimitry Andric default:
23bdd1243dSDimitry Andric return std::nullopt;
240b57cec5SDimitry Andric }
250b57cec5SDimitry Andric LLVM_BUILTIN_UNREACHABLE;
260b57cec5SDimitry Andric }
270b57cec5SDimitry Andric
consumeFieldLayout(StringRef & Spec,AlignStyle & Where,size_t & Align,char & Pad)280b57cec5SDimitry Andric bool formatv_object_base::consumeFieldLayout(StringRef &Spec, AlignStyle &Where,
290b57cec5SDimitry Andric size_t &Align, char &Pad) {
300b57cec5SDimitry Andric Where = AlignStyle::Right;
310b57cec5SDimitry Andric Align = 0;
320b57cec5SDimitry Andric Pad = ' ';
330b57cec5SDimitry Andric if (Spec.empty())
340b57cec5SDimitry Andric return true;
350b57cec5SDimitry Andric
360b57cec5SDimitry Andric if (Spec.size() > 1) {
370b57cec5SDimitry Andric // A maximum of 2 characters at the beginning can be used for something
380b57cec5SDimitry Andric // other
390b57cec5SDimitry Andric // than the width.
400b57cec5SDimitry Andric // If Spec[1] is a loc char, then Spec[0] is a pad char and Spec[2:...]
410b57cec5SDimitry Andric // contains the width.
420b57cec5SDimitry Andric // Otherwise, if Spec[0] is a loc char, then Spec[1:...] contains the width.
430b57cec5SDimitry Andric // Otherwise, Spec[0:...] contains the width.
440b57cec5SDimitry Andric if (auto Loc = translateLocChar(Spec[1])) {
450b57cec5SDimitry Andric Pad = Spec[0];
460b57cec5SDimitry Andric Where = *Loc;
470b57cec5SDimitry Andric Spec = Spec.drop_front(2);
480b57cec5SDimitry Andric } else if (auto Loc = translateLocChar(Spec[0])) {
490b57cec5SDimitry Andric Where = *Loc;
500b57cec5SDimitry Andric Spec = Spec.drop_front(1);
510b57cec5SDimitry Andric }
520b57cec5SDimitry Andric }
530b57cec5SDimitry Andric
540b57cec5SDimitry Andric bool Failed = Spec.consumeInteger(0, Align);
550b57cec5SDimitry Andric return !Failed;
560b57cec5SDimitry Andric }
570b57cec5SDimitry Andric
58bdd1243dSDimitry Andric std::optional<ReplacementItem>
parseReplacementItem(StringRef Spec)590b57cec5SDimitry Andric formatv_object_base::parseReplacementItem(StringRef Spec) {
600b57cec5SDimitry Andric StringRef RepString = Spec.trim("{}");
610b57cec5SDimitry Andric
620b57cec5SDimitry Andric // If the replacement sequence does not start with a non-negative integer,
630b57cec5SDimitry Andric // this is an error.
640b57cec5SDimitry Andric char Pad = ' ';
650b57cec5SDimitry Andric std::size_t Align = 0;
660b57cec5SDimitry Andric AlignStyle Where = AlignStyle::Right;
670b57cec5SDimitry Andric StringRef Options;
680b57cec5SDimitry Andric size_t Index = 0;
690b57cec5SDimitry Andric RepString = RepString.trim();
700b57cec5SDimitry Andric if (RepString.consumeInteger(0, Index)) {
710b57cec5SDimitry Andric assert(false && "Invalid replacement sequence index!");
720b57cec5SDimitry Andric return ReplacementItem{};
730b57cec5SDimitry Andric }
740b57cec5SDimitry Andric RepString = RepString.trim();
75*a58f00eaSDimitry Andric if (RepString.consume_front(",")) {
760b57cec5SDimitry Andric if (!consumeFieldLayout(RepString, Where, Align, Pad))
770b57cec5SDimitry Andric assert(false && "Invalid replacement field layout specification!");
780b57cec5SDimitry Andric }
790b57cec5SDimitry Andric RepString = RepString.trim();
800b57cec5SDimitry Andric if (!RepString.empty() && RepString.front() == ':') {
810b57cec5SDimitry Andric Options = RepString.drop_front().trim();
820b57cec5SDimitry Andric RepString = StringRef();
830b57cec5SDimitry Andric }
840b57cec5SDimitry Andric RepString = RepString.trim();
850b57cec5SDimitry Andric if (!RepString.empty()) {
860b57cec5SDimitry Andric assert(false && "Unexpected characters found in replacement string!");
870b57cec5SDimitry Andric }
880b57cec5SDimitry Andric
890b57cec5SDimitry Andric return ReplacementItem{Spec, Index, Align, Where, Pad, Options};
900b57cec5SDimitry Andric }
910b57cec5SDimitry Andric
920b57cec5SDimitry Andric std::pair<ReplacementItem, StringRef>
splitLiteralAndReplacement(StringRef Fmt)930b57cec5SDimitry Andric formatv_object_base::splitLiteralAndReplacement(StringRef Fmt) {
94e8d8bef9SDimitry Andric while (!Fmt.empty()) {
950b57cec5SDimitry Andric // Everything up until the first brace is a literal.
96e8d8bef9SDimitry Andric if (Fmt.front() != '{') {
97e8d8bef9SDimitry Andric std::size_t BO = Fmt.find_first_of('{');
980b57cec5SDimitry Andric return std::make_pair(ReplacementItem{Fmt.substr(0, BO)}, Fmt.substr(BO));
99e8d8bef9SDimitry Andric }
1000b57cec5SDimitry Andric
101e8d8bef9SDimitry Andric StringRef Braces = Fmt.take_while([](char C) { return C == '{'; });
1020b57cec5SDimitry Andric // If there is more than one brace, then some of them are escaped. Treat
1030b57cec5SDimitry Andric // these as replacements.
1040b57cec5SDimitry Andric if (Braces.size() > 1) {
1050b57cec5SDimitry Andric size_t NumEscapedBraces = Braces.size() / 2;
106e8d8bef9SDimitry Andric StringRef Middle = Fmt.take_front(NumEscapedBraces);
107e8d8bef9SDimitry Andric StringRef Right = Fmt.drop_front(NumEscapedBraces * 2);
1080b57cec5SDimitry Andric return std::make_pair(ReplacementItem{Middle}, Right);
1090b57cec5SDimitry Andric }
1100b57cec5SDimitry Andric // An unterminated open brace is undefined. We treat the rest of the string
1110b57cec5SDimitry Andric // as a literal replacement, but we assert to indicate that this is
1120b57cec5SDimitry Andric // undefined and that we consider it an error.
113e8d8bef9SDimitry Andric std::size_t BC = Fmt.find_first_of('}');
1140b57cec5SDimitry Andric if (BC == StringRef::npos) {
1150b57cec5SDimitry Andric assert(
1160b57cec5SDimitry Andric false &&
1170b57cec5SDimitry Andric "Unterminated brace sequence. Escape with {{ for a literal brace.");
1180b57cec5SDimitry Andric return std::make_pair(ReplacementItem{Fmt}, StringRef());
1190b57cec5SDimitry Andric }
1200b57cec5SDimitry Andric
1210b57cec5SDimitry Andric // Even if there is a closing brace, if there is another open brace before
1220b57cec5SDimitry Andric // this closing brace, treat this portion as literal, and try again with the
1230b57cec5SDimitry Andric // next one.
124e8d8bef9SDimitry Andric std::size_t BO2 = Fmt.find_first_of('{', 1);
1250b57cec5SDimitry Andric if (BO2 < BC)
1260b57cec5SDimitry Andric return std::make_pair(ReplacementItem{Fmt.substr(0, BO2)},
1270b57cec5SDimitry Andric Fmt.substr(BO2));
1280b57cec5SDimitry Andric
129e8d8bef9SDimitry Andric StringRef Spec = Fmt.slice(1, BC);
1300b57cec5SDimitry Andric StringRef Right = Fmt.substr(BC + 1);
1310b57cec5SDimitry Andric
1320b57cec5SDimitry Andric auto RI = parseReplacementItem(Spec);
13381ad6265SDimitry Andric if (RI)
1340b57cec5SDimitry Andric return std::make_pair(*RI, Right);
1350b57cec5SDimitry Andric
1360b57cec5SDimitry Andric // If there was an error parsing the replacement item, treat it as an
1370b57cec5SDimitry Andric // invalid replacement spec, and just continue.
138e8d8bef9SDimitry Andric Fmt = Fmt.drop_front(BC + 1);
1390b57cec5SDimitry Andric }
1400b57cec5SDimitry Andric return std::make_pair(ReplacementItem{Fmt}, StringRef());
1410b57cec5SDimitry Andric }
1420b57cec5SDimitry Andric
1435ffd83dbSDimitry Andric SmallVector<ReplacementItem, 2>
parseFormatString(StringRef Fmt)1440b57cec5SDimitry Andric formatv_object_base::parseFormatString(StringRef Fmt) {
1455ffd83dbSDimitry Andric SmallVector<ReplacementItem, 2> Replacements;
1460b57cec5SDimitry Andric ReplacementItem I;
1470b57cec5SDimitry Andric while (!Fmt.empty()) {
1480b57cec5SDimitry Andric std::tie(I, Fmt) = splitLiteralAndReplacement(Fmt);
1490b57cec5SDimitry Andric if (I.Type != ReplacementType::Empty)
1500b57cec5SDimitry Andric Replacements.push_back(I);
1510b57cec5SDimitry Andric }
1520b57cec5SDimitry Andric return Replacements;
1530b57cec5SDimitry Andric }
1540b57cec5SDimitry Andric
anchor()1550b57cec5SDimitry Andric void detail::format_adapter::anchor() { }
156