FileUtilities.cpp revision 360784 
1//===- Support/FileUtilities.cpp - File System Utilities ------------------===//
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 a family of utility functions which are useful for doing
10// various things with files.
11//
12//===----------------------------------------------------------------------===//
13
14#include "llvm/Support/FileUtilities.h"
15#include "llvm/ADT/ScopeExit.h"
16#include "llvm/ADT/SmallString.h"
17#include "llvm/Support/Error.h"
18#include "llvm/Support/ErrorOr.h"
19#include "llvm/Support/MemoryBuffer.h"
20#include "llvm/Support/Path.h"
21#include "llvm/Support/raw_ostream.h"
22#include <cctype>
23#include <cmath>
24#include <cstdint>
25#include <cstdlib>
26#include <cstring>
27#include <memory>
28#include <system_error>
29
30using namespace llvm;
31
32static bool isSignedChar(char C) {
33  return (C == '+' || C == '-');
34}
35
36static bool isExponentChar(char C) {
37  switch (C) {
38  case 'D':  // Strange exponential notation.
39  case 'd':  // Strange exponential notation.
40  case 'e':
41  case 'E': return true;
42  default: return false;
43  }
44}
45
46static bool isNumberChar(char C) {
47  switch (C) {
48  case '0': case '1': case '2': case '3': case '4':
49  case '5': case '6': case '7': case '8': case '9':
50  case '.': return true;
51  default: return isSignedChar(C) || isExponentChar(C);
52  }
53}
54
55static const char *BackupNumber(const char *Pos, const char *FirstChar) {
56  // If we didn't stop in the middle of a number, don't backup.
57  if (!isNumberChar(*Pos)) return Pos;
58
59  // Otherwise, return to the start of the number.
60  bool HasPeriod = false;
61  while (Pos > FirstChar && isNumberChar(Pos[-1])) {
62    // Backup over at most one period.
63    if (Pos[-1] == '.') {
64      if (HasPeriod)
65        break;
66      HasPeriod = true;
67    }
68
69    --Pos;
70    if (Pos > FirstChar && isSignedChar(Pos[0]) && !isExponentChar(Pos[-1]))
71      break;
72  }
73  return Pos;
74}
75
76/// EndOfNumber - Return the first character that is not part of the specified
77/// number.  This assumes that the buffer is null terminated, so it won't fall
78/// off the end.
79static const char *EndOfNumber(const char *Pos) {
80  while (isNumberChar(*Pos))
81    ++Pos;
82  return Pos;
83}
84
85/// CompareNumbers - compare two numbers, returning true if they are different.
86static bool CompareNumbers(const char *&F1P, const char *&F2P,
87                           const char *F1End, const char *F2End,
88                           double AbsTolerance, double RelTolerance,
89                           std::string *ErrorMsg) {
90  const char *F1NumEnd, *F2NumEnd;
91  double V1 = 0.0, V2 = 0.0;
92
93  // If one of the positions is at a space and the other isn't, chomp up 'til
94  // the end of the space.
95  while (isspace(static_cast<unsigned char>(*F1P)) && F1P != F1End)
96    ++F1P;
97  while (isspace(static_cast<unsigned char>(*F2P)) && F2P != F2End)
98    ++F2P;
99
100  // If we stop on numbers, compare their difference.
101  if (!isNumberChar(*F1P) || !isNumberChar(*F2P)) {
102    // The diff failed.
103    F1NumEnd = F1P;
104    F2NumEnd = F2P;
105  } else {
106    // Note that some ugliness is built into this to permit support for numbers
107    // that use "D" or "d" as their exponential marker, e.g. "1.234D45".  This
108    // occurs in 200.sixtrack in spec2k.
109    V1 = strtod(F1P, const_cast<char**>(&F1NumEnd));
110    V2 = strtod(F2P, const_cast<char**>(&F2NumEnd));
111
112    if (*F1NumEnd == 'D' || *F1NumEnd == 'd') {
113      // Copy string into tmp buffer to replace the 'D' with an 'e'.
114      SmallString<200> StrTmp(F1P, EndOfNumber(F1NumEnd)+1);
115      // Strange exponential notation!
116      StrTmp[static_cast<unsigned>(F1NumEnd-F1P)] = 'e';
117
118      V1 = strtod(&StrTmp[0], const_cast<char**>(&F1NumEnd));
119      F1NumEnd = F1P + (F1NumEnd-&StrTmp[0]);
120    }
121
122    if (*F2NumEnd == 'D' || *F2NumEnd == 'd') {
123      // Copy string into tmp buffer to replace the 'D' with an 'e'.
124      SmallString<200> StrTmp(F2P, EndOfNumber(F2NumEnd)+1);
125      // Strange exponential notation!
126      StrTmp[static_cast<unsigned>(F2NumEnd-F2P)] = 'e';
127
128      V2 = strtod(&StrTmp[0], const_cast<char**>(&F2NumEnd));
129      F2NumEnd = F2P + (F2NumEnd-&StrTmp[0]);
130    }
131  }
132
133  if (F1NumEnd == F1P || F2NumEnd == F2P) {
134    if (ErrorMsg) {
135      *ErrorMsg = "FP Comparison failed, not a numeric difference between '";
136      *ErrorMsg += F1P[0];
137      *ErrorMsg += "' and '";
138      *ErrorMsg += F2P[0];
139      *ErrorMsg += "'";
140    }
141    return true;
142  }
143
144  // Check to see if these are inside the absolute tolerance
145  if (AbsTolerance < std::abs(V1-V2)) {
146    // Nope, check the relative tolerance...
147    double Diff;
148    if (V2)
149      Diff = std::abs(V1/V2 - 1.0);
150    else if (V1)
151      Diff = std::abs(V2/V1 - 1.0);
152    else
153      Diff = 0;  // Both zero.
154    if (Diff > RelTolerance) {
155      if (ErrorMsg) {
156        raw_string_ostream(*ErrorMsg)
157          << "Compared: " << V1 << " and " << V2 << '\n'
158          << "abs. diff = " << std::abs(V1-V2) << " rel.diff = " << Diff << '\n'
159          << "Out of tolerance: rel/abs: " << RelTolerance << '/'
160          << AbsTolerance;
161      }
162      return true;
163    }
164  }
165
166  // Otherwise, advance our read pointers to the end of the numbers.
167  F1P = F1NumEnd;  F2P = F2NumEnd;
168  return false;
169}
170
171/// DiffFilesWithTolerance - Compare the two files specified, returning 0 if the
172/// files match, 1 if they are different, and 2 if there is a file error.  This
173/// function differs from DiffFiles in that you can specify an absolete and
174/// relative FP error that is allowed to exist.  If you specify a string to fill
175/// in for the error option, it will set the string to an error message if an
176/// error occurs, allowing the caller to distinguish between a failed diff and a
177/// file system error.
178///
179int llvm::DiffFilesWithTolerance(StringRef NameA,
180                                 StringRef NameB,
181                                 double AbsTol, double RelTol,
182                                 std::string *Error) {
183  // Now its safe to mmap the files into memory because both files
184  // have a non-zero size.
185  ErrorOr<std::unique_ptr<MemoryBuffer>> F1OrErr = MemoryBuffer::getFile(NameA);
186  if (std::error_code EC = F1OrErr.getError()) {
187    if (Error)
188      *Error = EC.message();
189    return 2;
190  }
191  MemoryBuffer &F1 = *F1OrErr.get();
192
193  ErrorOr<std::unique_ptr<MemoryBuffer>> F2OrErr = MemoryBuffer::getFile(NameB);
194  if (std::error_code EC = F2OrErr.getError()) {
195    if (Error)
196      *Error = EC.message();
197    return 2;
198  }
199  MemoryBuffer &F2 = *F2OrErr.get();
200
201  // Okay, now that we opened the files, scan them for the first difference.
202  const char *File1Start = F1.getBufferStart();
203  const char *File2Start = F2.getBufferStart();
204  const char *File1End = F1.getBufferEnd();
205  const char *File2End = F2.getBufferEnd();
206  const char *F1P = File1Start;
207  const char *F2P = File2Start;
208  uint64_t A_size = F1.getBufferSize();
209  uint64_t B_size = F2.getBufferSize();
210
211  // Are the buffers identical?  Common case: Handle this efficiently.
212  if (A_size == B_size &&
213      std::memcmp(File1Start, File2Start, A_size) == 0)
214    return 0;
215
216  // Otherwise, we are done a tolerances are set.
217  if (AbsTol == 0 && RelTol == 0) {
218    if (Error)
219      *Error = "Files differ without tolerance allowance";
220    return 1;   // Files different!
221  }
222
223  bool CompareFailed = false;
224  while (true) {
225    // Scan for the end of file or next difference.
226    while (F1P < File1End && F2P < File2End && *F1P == *F2P) {
227      ++F1P;
228      ++F2P;
229    }
230
231    if (F1P >= File1End || F2P >= File2End) break;
232
233    // Okay, we must have found a difference.  Backup to the start of the
234    // current number each stream is at so that we can compare from the
235    // beginning.
236    F1P = BackupNumber(F1P, File1Start);
237    F2P = BackupNumber(F2P, File2Start);
238
239    // Now that we are at the start of the numbers, compare them, exiting if
240    // they don't match.
241    if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error)) {
242      CompareFailed = true;
243      break;
244    }
245  }
246
247  // Okay, we reached the end of file.  If both files are at the end, we
248  // succeeded.
249  bool F1AtEnd = F1P >= File1End;
250  bool F2AtEnd = F2P >= File2End;
251  if (!CompareFailed && (!F1AtEnd || !F2AtEnd)) {
252    // Else, we might have run off the end due to a number: backup and retry.
253    if (F1AtEnd && isNumberChar(F1P[-1])) --F1P;
254    if (F2AtEnd && isNumberChar(F2P[-1])) --F2P;
255    F1P = BackupNumber(F1P, File1Start);
256    F2P = BackupNumber(F2P, File2Start);
257
258    // Now that we are at the start of the numbers, compare them, exiting if
259    // they don't match.
260    if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error))
261      CompareFailed = true;
262
263    // If we found the end, we succeeded.
264    if (F1P < File1End || F2P < File2End)
265      CompareFailed = true;
266  }
267
268  return CompareFailed;
269}
270
271void llvm::AtomicFileWriteError::log(raw_ostream &OS) const {
272  OS << "atomic_write_error: ";
273  switch (Error) {
274  case atomic_write_error::failed_to_create_uniq_file:
275    OS << "failed_to_create_uniq_file";
276    return;
277  case atomic_write_error::output_stream_error:
278    OS << "output_stream_error";
279    return;
280  case atomic_write_error::failed_to_rename_temp_file:
281    OS << "failed_to_rename_temp_file";
282    return;
283  }
284  llvm_unreachable("unknown atomic_write_error value in "
285                   "failed_to_rename_temp_file::log()");
286}
287
288llvm::Error llvm::writeFileAtomically(StringRef TempPathModel,
289                                      StringRef FinalPath, StringRef Buffer) {
290  return writeFileAtomically(TempPathModel, FinalPath,
291                             [&Buffer](llvm::raw_ostream &OS) {
292                               OS.write(Buffer.data(), Buffer.size());
293                               return llvm::Error::success();
294                             });
295}
296
297llvm::Error llvm::writeFileAtomically(
298    StringRef TempPathModel, StringRef FinalPath,
299    std::function<llvm::Error(llvm::raw_ostream &)> Writer) {
300  SmallString<128> GeneratedUniqPath;
301  int TempFD;
302  if (sys::fs::createUniqueFile(TempPathModel.str(), TempFD,
303                                GeneratedUniqPath)) {
304    return llvm::make_error<AtomicFileWriteError>(
305        atomic_write_error::failed_to_create_uniq_file);
306  }
307  llvm::FileRemover RemoveTmpFileOnFail(GeneratedUniqPath);
308
309  raw_fd_ostream OS(TempFD, /*shouldClose=*/true);
310  if (llvm::Error Err = Writer(OS)) {
311    return Err;
312  }
313
314  OS.close();
315  if (OS.has_error()) {
316    OS.clear_error();
317    return llvm::make_error<AtomicFileWriteError>(
318        atomic_write_error::output_stream_error);
319  }
320
321  if (const std::error_code Error =
322          sys::fs::rename(/*from=*/GeneratedUniqPath.c_str(),
323                          /*to=*/FinalPath.str().c_str())) {
324    return llvm::make_error<AtomicFileWriteError>(
325        atomic_write_error::failed_to_rename_temp_file);
326  }
327
328  RemoveTmpFileOnFail.releaseFile();
329  return Error::success();
330}
331
332char llvm::AtomicFileWriteError::ID;
333