LLVM  10.0.0svn
StringMatcher.cpp
Go to the documentation of this file.
1 //===- StringMatcher.cpp - Generate a matcher for input strings -----------===//
2 //
3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 // See https://llvm.org/LICENSE.txt for license information.
5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6 //
7 //===----------------------------------------------------------------------===//
8 //
9 // This file implements the StringMatcher class.
10 //
11 //===----------------------------------------------------------------------===//
12 
14 #include "llvm/ADT/StringRef.h"
16 #include <cassert>
17 #include <map>
18 #include <string>
19 #include <utility>
20 #include <vector>
21 
22 using namespace llvm;
23 
24 /// FindFirstNonCommonLetter - Find the first character in the keys of the
25 /// string pairs that is not shared across the whole set of strings. All
26 /// strings are assumed to have the same length.
27 static unsigned
28 FindFirstNonCommonLetter(const std::vector<const
29  StringMatcher::StringPair*> &Matches) {
30  assert(!Matches.empty());
31  for (unsigned i = 0, e = Matches[0]->first.size(); i != e; ++i) {
32  // Check to see if letter i is the same across the set.
33  char Letter = Matches[0]->first[i];
34 
35  for (unsigned str = 0, e = Matches.size(); str != e; ++str)
36  if (Matches[str]->first[i] != Letter)
37  return i;
38  }
39 
40  return Matches[0]->first.size();
41 }
42 
43 /// EmitStringMatcherForChar - Given a set of strings that are known to be the
44 /// same length and whose characters leading up to CharNo are the same, emit
45 /// code to verify that CharNo and later are the same.
46 ///
47 /// \return - True if control can leave the emitted code fragment.
48 bool StringMatcher::EmitStringMatcherForChar(
49  const std::vector<const StringPair *> &Matches, unsigned CharNo,
50  unsigned IndentCount, bool IgnoreDuplicates) const {
51  assert(!Matches.empty() && "Must have at least one string to match!");
52  std::string Indent(IndentCount * 2 + 4, ' ');
53 
54  // If we have verified that the entire string matches, we're done: output the
55  // matching code.
56  if (CharNo == Matches[0]->first.size()) {
57  if (Matches.size() > 1 && !IgnoreDuplicates)
58  report_fatal_error("Had duplicate keys to match on");
59 
60  // If the to-execute code has \n's in it, indent each subsequent line.
61  StringRef Code = Matches[0]->second;
62 
63  std::pair<StringRef, StringRef> Split = Code.split('\n');
64  OS << Indent << Split.first << "\t // \"" << Matches[0]->first << "\"\n";
65 
66  Code = Split.second;
67  while (!Code.empty()) {
68  Split = Code.split('\n');
69  OS << Indent << Split.first << "\n";
70  Code = Split.second;
71  }
72  return false;
73  }
74 
75  // Bucket the matches by the character we are comparing.
76  std::map<char, std::vector<const StringPair*>> MatchesByLetter;
77 
78  for (unsigned i = 0, e = Matches.size(); i != e; ++i)
79  MatchesByLetter[Matches[i]->first[CharNo]].push_back(Matches[i]);
80 
81 
82  // If we have exactly one bucket to match, see how many characters are common
83  // across the whole set and match all of them at once.
84  if (MatchesByLetter.size() == 1) {
85  unsigned FirstNonCommonLetter = FindFirstNonCommonLetter(Matches);
86  unsigned NumChars = FirstNonCommonLetter-CharNo;
87 
88  // Emit code to break out if the prefix doesn't match.
89  if (NumChars == 1) {
90  // Do the comparison with if (Str[1] != 'f')
91  // FIXME: Need to escape general characters.
92  OS << Indent << "if (" << StrVariableName << "[" << CharNo << "] != '"
93  << Matches[0]->first[CharNo] << "')\n";
94  OS << Indent << " break;\n";
95  } else {
96  // Do the comparison with if memcmp(Str.data()+1, "foo", 3).
97  // FIXME: Need to escape general strings.
98  OS << Indent << "if (memcmp(" << StrVariableName << ".data()+" << CharNo
99  << ", \"" << Matches[0]->first.substr(CharNo, NumChars) << "\", "
100  << NumChars << ") != 0)\n";
101  OS << Indent << " break;\n";
102  }
103 
104  return EmitStringMatcherForChar(Matches, FirstNonCommonLetter, IndentCount,
105  IgnoreDuplicates);
106  }
107 
108  // Otherwise, we have multiple possible things, emit a switch on the
109  // character.
110  OS << Indent << "switch (" << StrVariableName << "[" << CharNo << "]) {\n";
111  OS << Indent << "default: break;\n";
112 
113  for (std::map<char, std::vector<const StringPair*>>::iterator LI =
114  MatchesByLetter.begin(), E = MatchesByLetter.end(); LI != E; ++LI) {
115  // TODO: escape hard stuff (like \n) if we ever care about it.
116  OS << Indent << "case '" << LI->first << "':\t // "
117  << LI->second.size() << " string";
118  if (LI->second.size() != 1) OS << 's';
119  OS << " to match.\n";
120  if (EmitStringMatcherForChar(LI->second, CharNo + 1, IndentCount + 1,
121  IgnoreDuplicates))
122  OS << Indent << " break;\n";
123  }
124 
125  OS << Indent << "}\n";
126  return true;
127 }
128 
129 /// Emit - Top level entry point.
130 ///
131 void StringMatcher::Emit(unsigned Indent, bool IgnoreDuplicates) const {
132  // If nothing to match, just fall through.
133  if (Matches.empty()) return;
134 
135  // First level categorization: group strings by length.
136  std::map<unsigned, std::vector<const StringPair*>> MatchesByLength;
137 
138  for (unsigned i = 0, e = Matches.size(); i != e; ++i)
139  MatchesByLength[Matches[i].first.size()].push_back(&Matches[i]);
140 
141  // Output a switch statement on length and categorize the elements within each
142  // bin.
143  OS.indent(Indent*2+2) << "switch (" << StrVariableName << ".size()) {\n";
144  OS.indent(Indent*2+2) << "default: break;\n";
145 
146  for (std::map<unsigned, std::vector<const StringPair*>>::iterator LI =
147  MatchesByLength.begin(), E = MatchesByLength.end(); LI != E; ++LI) {
148  OS.indent(Indent*2+2) << "case " << LI->first << ":\t // "
149  << LI->second.size()
150  << " string" << (LI->second.size() == 1 ? "" : "s") << " to match.\n";
151  if (EmitStringMatcherForChar(LI->second, 0, Indent, IgnoreDuplicates))
152  OS.indent(Indent*2+4) << "break;\n";
153  }
154 
155  OS.indent(Indent*2+2) << "}\n";
156 }
LLVM_ATTRIBUTE_NORETURN void report_fatal_error(Error Err, bool gen_crash_diag=true)
Report a serious error, calling any installed error handler.
Definition: Error.cpp:139
This class represents lattice values for constants.
Definition: AllocatorList.h:23
raw_ostream & indent(unsigned NumSpaces)
indent - Insert &#39;NumSpaces&#39; spaces.
std::pair< std::string, std::string > StringPair
Definition: StringMatcher.h:32
LLVM_NODISCARD bool empty() const
empty - Check if the string is empty.
Definition: StringRef.h:140
static unsigned FindFirstNonCommonLetter(const std::vector< const StringMatcher::StringPair *> &Matches)
FindFirstNonCommonLetter - Find the first character in the keys of the string pairs that is not share...
static GCRegistry::Add< CoreCLRGC > E("coreclr", "CoreCLR-compatible GC")
unsigned first
LLVM_NODISCARD std::pair< StringRef, StringRef > split(char Separator) const
Split into two substrings around the first occurrence of a separator character.
Definition: StringRef.h:710
assert(ImpDefSCC.getReg()==AMDGPU::SCC &&ImpDefSCC.isDef())
void Emit(unsigned Indent=0, bool IgnoreDuplicates=false) const
Emit - Top level entry point.
StringRef - Represent a constant reference to a string, i.e.
Definition: StringRef.h:48