Logo ROOT  
Reference Guide
 
Loading...
Searching...
No Matches
RRawFile.cxx
Go to the documentation of this file.
1// @(#)root/io:$Id$
2// Author: Jakob Blomer
3
4/*************************************************************************
5 * Copyright (C) 1995-2018, Rene Brun and Fons Rademakers. *
6 * All rights reserved. *
7 * *
8 * For the licensing terms see $ROOTSYS/LICENSE. *
9 * For the list of contributors see $ROOTSYS/README/CREDITS. *
10 *************************************************************************/
11
12#include <ROOT/RConfig.h>
13#include <ROOT/RRawFile.hxx>
14#ifdef _WIN32
15#include <ROOT/RRawFileWin.hxx>
16#else
17#include <ROOT/RRawFileUnix.hxx>
18#endif
19
20#include "TError.h"
21#include "TPluginManager.h"
22#include "TROOT.h"
23
24#include <algorithm>
25#include <cctype> // for towlower
26#include <cerrno>
27#include <cstddef>
28#include <cstdint>
29#include <cstring>
30#include <stdexcept>
31#include <string>
32
33namespace {
34const char *kTransportSeparator = "://";
35// Corresponds to ELineBreaks
36#ifdef _WIN32
37const char *kLineBreakTokens[] = {"", "\r\n", "\n", "\r\n"};
38constexpr unsigned int kLineBreakTokenSizes[] = {0, 2, 1, 2};
39#else
40const char *kLineBreakTokens[] = {"", "\n", "\n", "\r\n"};
41constexpr unsigned int kLineBreakTokenSizes[] = {0, 1, 1, 2};
42#endif
43constexpr unsigned int kLineBuffer = 128; // On Readln, look for line-breaks in chunks of 128 bytes
44} // anonymous namespace
45
46size_t ROOT::Internal::RRawFile::RBlockBuffer::CopyTo(void *buffer, size_t nbytes, std::uint64_t offset)
47{
48 if (offset < fBufferOffset)
49 return 0;
50
51 size_t copiedBytes = 0;
52 std::uint64_t offsetInBuffer = offset - fBufferOffset;
53 if (offsetInBuffer < static_cast<std::uint64_t>(fBufferSize)) {
54 size_t bytesInBuffer = std::min(nbytes, static_cast<size_t>(fBufferSize - offsetInBuffer));
55 memcpy(buffer, fBuffer + offsetInBuffer, bytesInBuffer);
56 copiedBytes = bytesInBuffer;
57 }
58 return copiedBytes;
59}
60
61ROOT::Internal::RRawFile::RRawFile(std::string_view url, ROptions options)
63 fOptions(options), fFilePos(0)
64{
65}
66
68{
69 delete[] fBufferSpace;
70}
71
72std::unique_ptr<ROOT::Internal::RRawFile>
73ROOT::Internal::RRawFile::Create(std::string_view url, ROptions options)
74{
75 std::string transport = GetTransport(url);
76 if (transport == "file") {
77#ifdef _WIN32
78 return std::unique_ptr<RRawFile>(new RRawFileWin(url, options));
79#else
80 return std::unique_ptr<RRawFile>(new RRawFileUnix(url, options));
81#endif
82 }
83 if (transport == "http" || transport == "https" ||
84 transport == "root" || transport == "roots" ) {
85 std::string plgclass = transport.compare( 0, 4, "http" ) == 0 ?
86 "RRawFileDavix" : "RRawFileNetXNG";
87 if (TPluginHandler *h = gROOT->GetPluginManager()->
88 FindHandler("ROOT::Internal::RRawFile", std::string(url).c_str())) {
89 if (h->LoadPlugin() == 0) {
90 return std::unique_ptr<RRawFile>(reinterpret_cast<RRawFile *>(h->ExecPlugin(2, &url, &options)));
91 }
92 throw std::runtime_error("Cannot load plugin handler for " + plgclass);
93 }
94 throw std::runtime_error("Cannot find plugin handler for " + plgclass);
95 }
96 throw std::runtime_error("Unsupported transport protocol: " + transport);
97}
98
99void *ROOT::Internal::RRawFile::MapImpl(size_t /* nbytes */, std::uint64_t /* offset */,
100 std::uint64_t& /* mapdOffset */)
101{
102 throw std::runtime_error("Memory mapping unsupported");
103}
104
105void ROOT::Internal::RRawFile::ReadVImpl(RIOVec *ioVec, unsigned int nReq)
106{
107 for (unsigned i = 0; i < nReq; ++i) {
108 ioVec[i].fOutBytes = ReadAt(ioVec[i].fBuffer, ioVec[i].fSize, ioVec[i].fOffset);
109 }
110}
111
112void ROOT::Internal::RRawFile::UnmapImpl(void * /* region */, size_t /* nbytes */)
113{
114 throw std::runtime_error("Memory mapping unsupported");
115}
116
117std::string ROOT::Internal::RRawFile::GetLocation(std::string_view url)
118{
119 auto idx = url.find(kTransportSeparator);
120 if (idx == std::string_view::npos)
121 return std::string(url);
122 return std::string(url.substr(idx + strlen(kTransportSeparator)));
123}
124
126{
127 if (!fIsOpen)
128 OpenImpl();
129 fIsOpen = true;
130
131 if (fFileSize == kUnknownFileSize)
132 fFileSize = GetSizeImpl();
133 return fFileSize;
134}
135
137 return fUrl;
138}
139
140std::string ROOT::Internal::RRawFile::GetTransport(std::string_view url)
141{
142 auto idx = url.find(kTransportSeparator);
143 if (idx == std::string_view::npos)
144 return "file";
145 std::string transport(url.substr(0, idx));
146 std::transform(transport.begin(), transport.end(), transport.begin(), ::tolower);
147 return transport;
148}
149
150void *ROOT::Internal::RRawFile::Map(size_t nbytes, std::uint64_t offset, std::uint64_t &mapdOffset)
151{
152 if (!fIsOpen)
153 OpenImpl();
154 fIsOpen = true;
155 return MapImpl(nbytes, offset, mapdOffset);
156}
157
158size_t ROOT::Internal::RRawFile::Read(void *buffer, size_t nbytes)
159{
160 size_t res = ReadAt(buffer, nbytes, fFilePos);
161 fFilePos += res;
162 return res;
163}
164
165size_t ROOT::Internal::RRawFile::ReadAt(void *buffer, size_t nbytes, std::uint64_t offset)
166{
167 if (!fIsOpen)
168 OpenImpl();
169 R__ASSERT(fOptions.fBlockSize >= 0);
170 fIsOpen = true;
171
172 // "Large" reads are served directly, bypassing the cache
173 if (nbytes > static_cast<unsigned int>(fOptions.fBlockSize))
174 return ReadAtImpl(buffer, nbytes, offset);
175
176 if (fBufferSpace == nullptr) {
177 fBufferSpace = new unsigned char[kNumBlockBuffers * fOptions.fBlockSize];
178 for (unsigned int i = 0; i < kNumBlockBuffers; ++i)
179 fBlockBuffers[i].fBuffer = fBufferSpace + i * fOptions.fBlockSize;
180 }
181
182 size_t totalBytes = 0;
183 size_t copiedBytes = 0;
184 /// Try to serve as many bytes as possible from the block buffers
185 for (unsigned int idx = fBlockBufferIdx; idx < fBlockBufferIdx + kNumBlockBuffers; ++idx) {
186 copiedBytes = fBlockBuffers[idx % kNumBlockBuffers].CopyTo(buffer, nbytes, offset);
187 buffer = reinterpret_cast<unsigned char *>(buffer) + copiedBytes;
188 nbytes -= copiedBytes;
189 offset += copiedBytes;
190 totalBytes += copiedBytes;
191 if (copiedBytes > 0)
192 fBlockBufferIdx = idx;
193 if (nbytes == 0)
194 return totalBytes;
195 }
196 fBlockBufferIdx++;
197
198 /// The request was not fully satisfied and fBlockBufferIdx now points to the previous shadow buffer
199
200 /// The remaining bytes populate the newly promoted main buffer
201 RBlockBuffer *thisBuffer = &fBlockBuffers[fBlockBufferIdx % kNumBlockBuffers];
202 size_t res = ReadAtImpl(thisBuffer->fBuffer, fOptions.fBlockSize, offset);
203 thisBuffer->fBufferOffset = offset;
204 thisBuffer->fBufferSize = res;
205 size_t remainingBytes = std::min(res, nbytes);
206 memcpy(buffer, thisBuffer->fBuffer, remainingBytes);
207 totalBytes += remainingBytes;
208 return totalBytes;
209}
210
211void ROOT::Internal::RRawFile::ReadV(RIOVec *ioVec, unsigned int nReq)
212{
213 if (!fIsOpen)
214 OpenImpl();
215 fIsOpen = true;
216 ReadVImpl(ioVec, nReq);
217}
218
220{
221 if (fOptions.fLineBreak == ELineBreaks::kAuto) {
222 // Auto-detect line breaks according to the break discovered in the first line
223 fOptions.fLineBreak = ELineBreaks::kUnix;
224 bool res = Readln(line);
225 if ((line.length() > 0) && (*line.rbegin() == '\r')) {
226 fOptions.fLineBreak = ELineBreaks::kWindows;
227 line.resize(line.length() - 1);
228 }
229 return res;
230 }
231
232 line.clear();
233 char buffer[kLineBuffer];
234 size_t nbytes;
235 do {
236 nbytes = Read(buffer, sizeof(buffer));
237 std::string_view bufferView(buffer, nbytes);
238 auto idx = bufferView.find(kLineBreakTokens[static_cast<int>(fOptions.fLineBreak)]);
239 if (idx != std::string_view::npos) {
240 // Line break found, return the string and skip the linebreak itself
241 line.append(buffer, idx);
242 fFilePos -= nbytes - idx;
243 fFilePos += kLineBreakTokenSizes[static_cast<int>(fOptions.fLineBreak)];
244 return true;
245 }
246 line.append(buffer, nbytes);
247 } while (nbytes > 0);
248
249 return !line.empty();
250}
251
252void ROOT::Internal::RRawFile::Seek(std::uint64_t offset)
253{
254 fFilePos = offset;
255}
256
257void ROOT::Internal::RRawFile::Unmap(void *region, size_t nbytes)
258{
259 if (!fIsOpen)
260 throw std::runtime_error("Cannot unmap, file not open");
261 UnmapImpl(region, nbytes);
262}
size_t fSize
std::string fBuffer
#define h(i)
Definition RSha256.hxx:106
#define R__ASSERT(e)
Definition TError.h:118
#define gROOT
Definition TROOT.h:404
The RRawFileUnix class uses POSIX calls to read from a mounted file system.
The RRawFileWin class uses portable C I/O calls to read from a drive.
The RRawFile provides read-only access to local and remote files.
Definition RRawFile.hxx:43
unsigned char * fBufferSpace
Memory block containing the block buffers consecutively.
Definition RRawFile.hxx:106
static std::string GetLocation(std::string_view url)
Returns only the file location, e.g. "server/file" for http://server/file.
Definition RRawFile.cxx:117
unsigned int fBlockBufferIdx
To be used modulo kNumBlockBuffers, points to the last used block buffer in fBlockBuffers.
Definition RRawFile.hxx:102
RRawFile(std::string_view url, ROptions options)
Definition RRawFile.cxx:61
std::uint64_t fFilePos
The current position in the file, which can be changed by Seek, Read, and Readln.
Definition RRawFile.hxx:116
virtual void * MapImpl(size_t nbytes, std::uint64_t offset, std::uint64_t &mapdOffset)
If a derived class supports mmap, the MapImpl and UnmapImpl calls are supposed to be implemented,...
Definition RRawFile.cxx:99
void Unmap(void *region, size_t nbytes)
Receives a pointer returned by Map() and should have nbytes set to the full length of the mapping.
Definition RRawFile.cxx:257
virtual void ReadVImpl(RIOVec *ioVec, unsigned int nReq)
By default implemented as a loop of ReadAt calls but can be overwritten, e.g. XRootD or DAVIX impleme...
Definition RRawFile.cxx:105
static std::string GetTransport(std::string_view url)
Returns only the transport protocol in lower case, e.g. "http" for HTTP://server/file.
Definition RRawFile.cxx:140
std::uint64_t GetSize()
Returns the size of the file.
Definition RRawFile.cxx:125
static std::unique_ptr< RRawFile > Create(std::string_view url, ROptions options=ROptions())
Factory method that returns a suitable concrete implementation according to the transport in the url.
Definition RRawFile.cxx:73
void Seek(std::uint64_t offset)
Change the cursor fFilePos.
Definition RRawFile.cxx:252
static constexpr std::uint64_t kUnknownFileSize
Derived classes do not necessarily need to provide file size information but they can return "not kno...
Definition RRawFile.hxx:46
size_t ReadAt(void *buffer, size_t nbytes, std::uint64_t offset)
Buffered read from a random position.
Definition RRawFile.cxx:165
bool fIsOpen
Files are opened lazily and only when required; the open state is kept by this flag.
Definition RRawFile.hxx:110
bool Readln(std::string &line)
Read the next line starting from the current value of fFilePos. Returns false if the end of the file ...
Definition RRawFile.cxx:219
void ReadV(RIOVec *ioVec, unsigned int nReq)
Opens the file if necessary and calls ReadVImpl.
Definition RRawFile.cxx:211
size_t Read(void *buffer, size_t nbytes)
Read from fFilePos offset. Returns the actual number of bytes read.
Definition RRawFile.cxx:158
std::string GetUrl() const
Returns the url of the file.
Definition RRawFile.cxx:136
std::uint64_t fFileSize
The cached file size.
Definition RRawFile.hxx:108
virtual void UnmapImpl(void *region, size_t nbytes)
Derived classes with mmap support must be able to unmap the memory area handed out by Map()
Definition RRawFile.cxx:112
void * Map(size_t nbytes, std::uint64_t offset, std::uint64_t &mapdOffset)
Memory mapping according to POSIX standard; in particular, new mappings of the same range replace old...
Definition RRawFile.cxx:150
TLine * line
std::uint64_t fBufferOffset
Where in the open file does fBuffer start.
Definition RRawFile.hxx:87
unsigned char * fBuffer
Points into the I/O buffer with data from the file, not owned.
Definition RRawFile.hxx:91
size_t CopyTo(void *buffer, size_t nbytes, std::uint64_t offset)
Tries to copy up to nbytes starting at offset from fBuffer into buffer. Returns number of bytes copie...
Definition RRawFile.cxx:46
size_t fBufferSize
The number of currently buffered bytes in fBuffer.
Definition RRawFile.hxx:89
Used for vector reads from multiple offsets into multiple buffers.
Definition RRawFile.hxx:71
std::size_t fOutBytes
The number of actually read bytes, set by ReadV()
Definition RRawFile.hxx:79
On construction, an ROptions parameter can customize the RRawFile behavior.
Definition RRawFile.hxx:59