From 44c35e80b165cbd193a39966b7ba434b676071c0 Mon Sep 17 00:00:00 2001 From: Zachary Turner Date: Mon, 29 Aug 2016 19:45:59 +0000 Subject: [PATCH] Copy StringExtractor to StdStringExtractor. I have some improvements to make to StringExtractor that require using LLVM. debugserver can't take a dependency on LLVM but uses this file, so I'm forking it off into StdStringExtractor and StringExtractor, so that StringExtractor can take advantage of some performance improvements and readability improvements that LLVM can provide. llvm-svn: 279997 --- lldb/include/lldb/Utility/StdStringExtractor.h | 181 +++++++++ lldb/source/Utility/CMakeLists.txt | 1 + lldb/source/Utility/StdStringExtractor.cpp | 502 +++++++++++++++++++++++++ lldb/tools/debugserver/source/JSON.cpp | 2 +- lldb/tools/debugserver/source/JSON.h | 6 +- lldb/tools/debugserver/source/RNBRemote.cpp | 24 +- 6 files changed, 700 insertions(+), 16 deletions(-) create mode 100644 lldb/include/lldb/Utility/StdStringExtractor.h create mode 100644 lldb/source/Utility/StdStringExtractor.cpp diff --git a/lldb/include/lldb/Utility/StdStringExtractor.h b/lldb/include/lldb/Utility/StdStringExtractor.h new file mode 100644 index 0000000..27229ff --- /dev/null +++ b/lldb/include/lldb/Utility/StdStringExtractor.h @@ -0,0 +1,181 @@ +//===-- StdStringExtractor.h ------------------------------------*- C++ -*-===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#ifndef utility_StdStringExtractor_h_ +#define utility_StdStringExtractor_h_ + +// C Includes +// C++ Includes +#include +#include + +// Other libraries and framework includes +// Project includes + +// Based on StringExtractor, with the added limitation that this file should not +// take a dependency on LLVM, as it is used from debugserver. +class StdStringExtractor +{ +public: + + enum { + BigEndian = 0, + LittleEndian = 1 + }; + //------------------------------------------------------------------ + // Constructors and Destructors + //------------------------------------------------------------------ + StdStringExtractor(); + StdStringExtractor(const char *packet_cstr); + StdStringExtractor(const StdStringExtractor& rhs); + virtual ~StdStringExtractor(); + + //------------------------------------------------------------------ + // Operators + //------------------------------------------------------------------ + const StdStringExtractor& + operator=(const StdStringExtractor& rhs); + + // Returns true if the file position is still valid for the data + // contained in this string extractor object. + bool + IsGood() const + { + return m_index != UINT64_MAX; + } + + uint64_t + GetFilePos () const + { + return m_index; + } + + void + SetFilePos (uint32_t idx) + { + m_index = idx; + } + + void + Clear () + { + m_packet.clear(); + m_index = 0; + } + + void + SkipSpaces (); + + std::string & + GetStringRef () + { + return m_packet; + } + + const std::string & + GetStringRef () const + { + return m_packet; + } + + bool + Empty() + { + return m_packet.empty(); + } + + size_t + GetBytesLeft () + { + if (m_index < m_packet.size()) + return m_packet.size() - m_index; + return 0; + } + + char + GetChar (char fail_value = '\0'); + + char + PeekChar (char fail_value = '\0') + { + const char *cstr = Peek(); + if (cstr) + return cstr[0]; + return fail_value; + } + + int + DecodeHexU8(); + + uint8_t + GetHexU8 (uint8_t fail_value = 0, bool set_eof_on_fail = true); + + bool + GetHexU8Ex (uint8_t& ch, bool set_eof_on_fail = true); + + bool + GetNameColonValue (std::string &name, std::string &value); + + int32_t + GetS32 (int32_t fail_value, int base = 0); + + uint32_t + GetU32 (uint32_t fail_value, int base = 0); + + int64_t + GetS64 (int64_t fail_value, int base = 0); + + uint64_t + GetU64 (uint64_t fail_value, int base = 0); + + uint32_t + GetHexMaxU32 (bool little_endian, uint32_t fail_value); + + uint64_t + GetHexMaxU64 (bool little_endian, uint64_t fail_value); + + size_t + GetHexBytes (void *dst, size_t dst_len, uint8_t fail_fill_value); + + size_t + GetHexBytesAvail (void *dst, size_t dst_len); + + uint64_t + GetHexWithFixedSize (uint32_t byte_size, bool little_endian, uint64_t fail_value); + + size_t + GetHexByteString (std::string &str); + + size_t + GetHexByteStringFixedLength (std::string &str, uint32_t nibble_length); + + size_t + GetHexByteStringTerminatedBy (std::string &str, + char terminator); + + const char * + Peek () + { + if (m_index < m_packet.size()) + return m_packet.c_str() + m_index; + return nullptr; + } + +protected: + //------------------------------------------------------------------ + // For StdStringExtractor only + //------------------------------------------------------------------ + std::string m_packet; // The string in which to extract data. + uint64_t m_index; // When extracting data from a packet, this index + // will march along as things get extracted. If set + // to UINT64_MAX the end of the packet data was + // reached when decoding information +}; + +#endif // utility_StringExtractor_h_ diff --git a/lldb/source/Utility/CMakeLists.txt b/lldb/source/Utility/CMakeLists.txt index 90aff22..72c38b2 100644 --- a/lldb/source/Utility/CMakeLists.txt +++ b/lldb/source/Utility/CMakeLists.txt @@ -12,6 +12,7 @@ add_lldb_library(lldbUtility RegisterNumber.cpp SelectHelper.cpp SharingPtr.cpp + StdStringExtractor.cpp StringExtractor.cpp StringExtractorGDBRemote.cpp StringLexer.cpp diff --git a/lldb/source/Utility/StdStringExtractor.cpp b/lldb/source/Utility/StdStringExtractor.cpp new file mode 100644 index 0000000..5152f3e2 --- /dev/null +++ b/lldb/source/Utility/StdStringExtractor.cpp @@ -0,0 +1,502 @@ +//===-- StdStringExtractor.cpp ----------------------------------*- C++ -*-===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#include "lldb/Utility/StdStringExtractor.h" + +// C Includes +#include + +// C++ Includes +// Other libraries and framework includes +// Project includes + +static inline int +xdigit_to_sint (char ch) +{ + if (ch >= 'a' && ch <= 'f') + return 10 + ch - 'a'; + if (ch >= 'A' && ch <= 'F') + return 10 + ch - 'A'; + if (ch >= '0' && ch <= '9') + return ch - '0'; + return -1; +} + +//---------------------------------------------------------------------- +// StdStringExtractor constructor +//---------------------------------------------------------------------- +StdStringExtractor::StdStringExtractor() : + m_packet(), + m_index (0) +{ +} + + +StdStringExtractor::StdStringExtractor(const char *packet_cstr) : + m_packet(), + m_index (0) +{ + if (packet_cstr) + m_packet.assign (packet_cstr); +} + + +//---------------------------------------------------------------------- +// StdStringExtractor copy constructor +//---------------------------------------------------------------------- +StdStringExtractor::StdStringExtractor(const StdStringExtractor& rhs) : + m_packet (rhs.m_packet), + m_index (rhs.m_index) +{ + +} + +//---------------------------------------------------------------------- +// StdStringExtractor assignment operator +//---------------------------------------------------------------------- +const StdStringExtractor& +StdStringExtractor::operator=(const StdStringExtractor& rhs) +{ + if (this != &rhs) + { + m_packet = rhs.m_packet; + m_index = rhs.m_index; + + } + return *this; +} + +//---------------------------------------------------------------------- +// Destructor +//---------------------------------------------------------------------- +StdStringExtractor::~StdStringExtractor() +{ +} + + +char +StdStringExtractor::GetChar (char fail_value) +{ + if (m_index < m_packet.size()) + { + char ch = m_packet[m_index]; + ++m_index; + return ch; + } + m_index = UINT64_MAX; + return fail_value; +} + +//---------------------------------------------------------------------- +// If a pair of valid hex digits exist at the head of the +// StdStringExtractor they are decoded into an unsigned byte and returned +// by this function +// +// If there is not a pair of valid hex digits at the head of the +// StdStringExtractor, it is left unchanged and -1 is returned +//---------------------------------------------------------------------- +int +StdStringExtractor::DecodeHexU8() +{ + SkipSpaces(); + if (GetBytesLeft() < 2) + { + return -1; + } + const int hi_nibble = xdigit_to_sint(m_packet[m_index]); + const int lo_nibble = xdigit_to_sint(m_packet[m_index+1]); + if (hi_nibble == -1 || lo_nibble == -1) + { + return -1; + } + m_index += 2; + return (uint8_t)((hi_nibble << 4) + lo_nibble); +} + +//---------------------------------------------------------------------- +// Extract an unsigned character from two hex ASCII chars in the packet +// string, or return fail_value on failure +//---------------------------------------------------------------------- +uint8_t +StdStringExtractor::GetHexU8 (uint8_t fail_value, bool set_eof_on_fail) +{ + // On success, fail_value will be overwritten with the next + // character in the stream + GetHexU8Ex(fail_value, set_eof_on_fail); + return fail_value; +} + +bool +StdStringExtractor::GetHexU8Ex (uint8_t& ch, bool set_eof_on_fail) +{ + int byte = DecodeHexU8(); + if (byte == -1) + { + if (set_eof_on_fail || m_index >= m_packet.size()) + m_index = UINT64_MAX; + // ch should not be changed in case of failure + return false; + } + ch = (uint8_t)byte; + return true; +} + +uint32_t +StdStringExtractor::GetU32 (uint32_t fail_value, int base) +{ + if (m_index < m_packet.size()) + { + char *end = nullptr; + const char *start = m_packet.c_str(); + const char *cstr = start + m_index; + uint32_t result = static_cast(::strtoul (cstr, &end, base)); + + if (end && end != cstr) + { + m_index = end - start; + return result; + } + } + return fail_value; +} + +int32_t +StdStringExtractor::GetS32 (int32_t fail_value, int base) +{ + if (m_index < m_packet.size()) + { + char *end = nullptr; + const char *start = m_packet.c_str(); + const char *cstr = start + m_index; + int32_t result = static_cast(::strtol (cstr, &end, base)); + + if (end && end != cstr) + { + m_index = end - start; + return result; + } + } + return fail_value; +} + + +uint64_t +StdStringExtractor::GetU64 (uint64_t fail_value, int base) +{ + if (m_index < m_packet.size()) + { + char *end = nullptr; + const char *start = m_packet.c_str(); + const char *cstr = start + m_index; + uint64_t result = ::strtoull (cstr, &end, base); + + if (end && end != cstr) + { + m_index = end - start; + return result; + } + } + return fail_value; +} + +int64_t +StdStringExtractor::GetS64 (int64_t fail_value, int base) +{ + if (m_index < m_packet.size()) + { + char *end = nullptr; + const char *start = m_packet.c_str(); + const char *cstr = start + m_index; + int64_t result = ::strtoll (cstr, &end, base); + + if (end && end != cstr) + { + m_index = end - start; + return result; + } + } + return fail_value; +} + + +uint32_t +StdStringExtractor::GetHexMaxU32 (bool little_endian, uint32_t fail_value) +{ + uint32_t result = 0; + uint32_t nibble_count = 0; + + SkipSpaces(); + if (little_endian) + { + uint32_t shift_amount = 0; + while (m_index < m_packet.size() && ::isxdigit (m_packet[m_index])) + { + // Make sure we don't exceed the size of a uint32_t... + if (nibble_count >= (sizeof(uint32_t) * 2)) + { + m_index = UINT64_MAX; + return fail_value; + } + + uint8_t nibble_lo; + uint8_t nibble_hi = xdigit_to_sint (m_packet[m_index]); + ++m_index; + if (m_index < m_packet.size() && ::isxdigit (m_packet[m_index])) + { + nibble_lo = xdigit_to_sint (m_packet[m_index]); + ++m_index; + result |= ((uint32_t)nibble_hi << (shift_amount + 4)); + result |= ((uint32_t)nibble_lo << shift_amount); + nibble_count += 2; + shift_amount += 8; + } + else + { + result |= ((uint32_t)nibble_hi << shift_amount); + nibble_count += 1; + shift_amount += 4; + } + + } + } + else + { + while (m_index < m_packet.size() && ::isxdigit (m_packet[m_index])) + { + // Make sure we don't exceed the size of a uint32_t... + if (nibble_count >= (sizeof(uint32_t) * 2)) + { + m_index = UINT64_MAX; + return fail_value; + } + + uint8_t nibble = xdigit_to_sint (m_packet[m_index]); + // Big Endian + result <<= 4; + result |= nibble; + + ++m_index; + ++nibble_count; + } + } + return result; +} + +uint64_t +StdStringExtractor::GetHexMaxU64 (bool little_endian, uint64_t fail_value) +{ + uint64_t result = 0; + uint32_t nibble_count = 0; + + SkipSpaces(); + if (little_endian) + { + uint32_t shift_amount = 0; + while (m_index < m_packet.size() && ::isxdigit (m_packet[m_index])) + { + // Make sure we don't exceed the size of a uint64_t... + if (nibble_count >= (sizeof(uint64_t) * 2)) + { + m_index = UINT64_MAX; + return fail_value; + } + + uint8_t nibble_lo; + uint8_t nibble_hi = xdigit_to_sint (m_packet[m_index]); + ++m_index; + if (m_index < m_packet.size() && ::isxdigit (m_packet[m_index])) + { + nibble_lo = xdigit_to_sint (m_packet[m_index]); + ++m_index; + result |= ((uint64_t)nibble_hi << (shift_amount + 4)); + result |= ((uint64_t)nibble_lo << shift_amount); + nibble_count += 2; + shift_amount += 8; + } + else + { + result |= ((uint64_t)nibble_hi << shift_amount); + nibble_count += 1; + shift_amount += 4; + } + + } + } + else + { + while (m_index < m_packet.size() && ::isxdigit (m_packet[m_index])) + { + // Make sure we don't exceed the size of a uint64_t... + if (nibble_count >= (sizeof(uint64_t) * 2)) + { + m_index = UINT64_MAX; + return fail_value; + } + + uint8_t nibble = xdigit_to_sint (m_packet[m_index]); + // Big Endian + result <<= 4; + result |= nibble; + + ++m_index; + ++nibble_count; + } + } + return result; +} + +size_t +StdStringExtractor::GetHexBytes (void *dst_void, size_t dst_len, uint8_t fail_fill_value) +{ + uint8_t *dst = (uint8_t*)dst_void; + size_t bytes_extracted = 0; + while (bytes_extracted < dst_len && GetBytesLeft ()) + { + dst[bytes_extracted] = GetHexU8 (fail_fill_value); + if (IsGood()) + ++bytes_extracted; + else + break; + } + + for (size_t i = bytes_extracted; i < dst_len; ++i) + dst[i] = fail_fill_value; + + return bytes_extracted; +} + +//---------------------------------------------------------------------- +// Decodes all valid hex encoded bytes at the head of the +// StdStringExtractor, limited by dst_len. +// +// Returns the number of bytes successfully decoded +//---------------------------------------------------------------------- +size_t +StdStringExtractor::GetHexBytesAvail (void *dst_void, size_t dst_len) +{ + uint8_t *dst = (uint8_t*)dst_void; + size_t bytes_extracted = 0; + while (bytes_extracted < dst_len) + { + int decode = DecodeHexU8(); + if (decode == -1) + { + break; + } + dst[bytes_extracted++] = (uint8_t)decode; + } + return bytes_extracted; +} + +// Consume ASCII hex nibble character pairs until we have decoded byte_size +// bytes of data. + +uint64_t +StdStringExtractor::GetHexWithFixedSize (uint32_t byte_size, bool little_endian, uint64_t fail_value) +{ + if (byte_size <= 8 && GetBytesLeft() >= byte_size * 2) + { + uint64_t result = 0; + uint32_t i; + if (little_endian) + { + // Little Endian + uint32_t shift_amount; + for (i = 0, shift_amount = 0; + i < byte_size && IsGood(); + ++i, shift_amount += 8) + { + result |= ((uint64_t)GetHexU8() << shift_amount); + } + } + else + { + // Big Endian + for (i = 0; i < byte_size && IsGood(); ++i) + { + result <<= 8; + result |= GetHexU8(); + } + } + } + m_index = UINT64_MAX; + return fail_value; +} + +size_t +StdStringExtractor::GetHexByteString (std::string &str) +{ + str.clear(); + str.reserve(GetBytesLeft() / 2); + char ch; + while ((ch = GetHexU8()) != '\0') + str.append(1, ch); + return str.size(); +} + +size_t +StdStringExtractor::GetHexByteStringFixedLength (std::string &str, uint32_t nibble_length) +{ + str.clear(); + + uint32_t nibble_count = 0; + for (const char *pch = Peek(); (nibble_count < nibble_length) && (pch != nullptr); str.append(1, GetHexU8(0, false)), pch = Peek (), nibble_count += 2) + {} + + return str.size(); +} + +size_t +StdStringExtractor::GetHexByteStringTerminatedBy (std::string &str, + char terminator) +{ + str.clear(); + char ch; + while ((ch = GetHexU8(0,false)) != '\0') + str.append(1, ch); + if (Peek() && *Peek() == terminator) + return str.size(); + + str.clear(); + return str.size(); +} + +bool +StdStringExtractor::GetNameColonValue (std::string &name, std::string &value) +{ + // Read something in the form of NNNN:VVVV; where NNNN is any character + // that is not a colon, followed by a ':' character, then a value (one or + // more ';' chars), followed by a ';' + if (m_index < m_packet.size()) + { + const size_t colon_idx = m_packet.find (':', m_index); + if (colon_idx != std::string::npos) + { + const size_t semicolon_idx = m_packet.find (';', colon_idx); + if (semicolon_idx != std::string::npos) + { + name.assign (m_packet, m_index, colon_idx - m_index); + value.assign (m_packet, colon_idx + 1, semicolon_idx - (colon_idx + 1)); + m_index = semicolon_idx + 1; + return true; + } + } + } + m_index = UINT64_MAX; + return false; +} + +void +StdStringExtractor::SkipSpaces () +{ + const size_t n = m_packet.size(); + while (m_index < n && isspace(m_packet[m_index])) + ++m_index; +} + diff --git a/lldb/tools/debugserver/source/JSON.cpp b/lldb/tools/debugserver/source/JSON.cpp index 9d56d19..34a01f4 100644 --- a/lldb/tools/debugserver/source/JSON.cpp +++ b/lldb/tools/debugserver/source/JSON.cpp @@ -319,7 +319,7 @@ JSONArray::GetNumElements () JSONParser::JSONParser (const char *cstr) : - StringExtractor(cstr) + StdStringExtractor(cstr) { } diff --git a/lldb/tools/debugserver/source/JSON.h b/lldb/tools/debugserver/source/JSON.h index d1723d8..e24f928 100644 --- a/lldb/tools/debugserver/source/JSON.h +++ b/lldb/tools/debugserver/source/JSON.h @@ -10,9 +10,9 @@ #ifndef utility_JSON_h_ #define utility_JSON_h_ -// This cross-project usage is fine as StringExtractor.h is entirely +// This cross-project usage is fine as StdStringExtractor.h is entirely // self-contained. -#include "lldb/Utility/StringExtractor.h" +#include "lldb/Utility/StdStringExtractor.h" // C includes #include @@ -338,7 +338,7 @@ public: Vector m_elements; }; -class JSONParser : public StringExtractor +class JSONParser : public StdStringExtractor { public: enum Token diff --git a/lldb/tools/debugserver/source/RNBRemote.cpp b/lldb/tools/debugserver/source/RNBRemote.cpp index 11abe3c..e737913 100644 --- a/lldb/tools/debugserver/source/RNBRemote.cpp +++ b/lldb/tools/debugserver/source/RNBRemote.cpp @@ -38,7 +38,7 @@ #include "RNBServices.h" #include "RNBSocket.h" #include "JSON.h" -#include "lldb/Utility/StringExtractor.h" +#include "lldb/Utility/StdStringExtractor.h" #include "MacOSX/Genealogy.h" #include "JSONGenerator.h" @@ -2329,7 +2329,7 @@ RNBRemote::HandlePacket_QSetSTDIO (const char *p) // QSetSTDIN // QSetSTDOUT // QSetSTDERR - StringExtractor packet(p); + StdStringExtractor packet(p); packet.SetFilePos (7); char ch = packet.GetChar(); while (packet.GetChar() != ':') @@ -2368,7 +2368,7 @@ RNBRemote::HandlePacket_QSetWorkingDir (const char *p) // Only set the working directory if we don't already have a process if (!m_ctx.HasValidProcessID()) { - StringExtractor packet(p += sizeof ("QSetWorkingDir:") - 1); + StdStringExtractor packet(p += sizeof ("QSetWorkingDir:") - 1); if (packet.GetHexByteString (m_ctx.GetWorkingDir())) { struct stat working_dir_stat; @@ -3473,7 +3473,7 @@ RNBRemote::HandlePacket_G (const char *p) if (g_num_reg_entries == 0) InitializeRegisters (); - StringExtractor packet(p); + StdStringExtractor packet(p); packet.SetFilePos(1); // Skip the 'G' nub_process_t pid = m_ctx.ProcessID(); @@ -3540,10 +3540,10 @@ RNBRemoteShouldCancelCallback (void *not_used) rnb_err_t RNBRemote::HandlePacket_AllocateMemory (const char *p) { - StringExtractor packet (p); + StdStringExtractor packet (p); packet.SetFilePos(2); // Skip the "_M" - nub_addr_t size = packet.GetHexMaxU64 (StringExtractor::BigEndian, 0); + nub_addr_t size = packet.GetHexMaxU64 (StdStringExtractor::BigEndian, 0); if (size != 0) { if (packet.GetChar() == ',') @@ -3590,9 +3590,9 @@ RNBRemote::HandlePacket_AllocateMemory (const char *p) rnb_err_t RNBRemote::HandlePacket_DeallocateMemory (const char *p) { - StringExtractor packet (p); + StdStringExtractor packet (p); packet.SetFilePos(2); // Skip the "_m" - nub_addr_t addr = packet.GetHexMaxU64 (StringExtractor::BigEndian, INVALID_NUB_ADDRESS); + nub_addr_t addr = packet.GetHexMaxU64 (StdStringExtractor::BigEndian, INVALID_NUB_ADDRESS); if (addr != INVALID_NUB_ADDRESS) { @@ -3669,7 +3669,7 @@ RNBRemote::HandlePacket_RestoreRegisterState (const char *p) return HandlePacket_ILLFORMED (__FILE__, __LINE__, p, "No thread was is set with the Hg packet"); } - StringExtractor packet (p); + StdStringExtractor packet (p); packet.SetFilePos(strlen("QRestoreRegisterState:")); // Skip the "QRestoreRegisterState:" const uint32_t save_id = packet.GetU32(0); @@ -4222,7 +4222,7 @@ RNBRemote::HandlePacket_P (const char *p) nub_process_t pid = m_ctx.ProcessID(); - StringExtractor packet (p); + StdStringExtractor packet (p); const char cmd_char = packet.GetChar(); // Register ID is always in big endian @@ -4372,7 +4372,7 @@ RNBRemote::HandlePacket_GetProfileData (const char *p) if (pid == INVALID_NUB_PROCESS) return SendPacket ("OK"); - StringExtractor packet(p += sizeof ("qGetProfileData")); + StdStringExtractor packet(p += sizeof ("qGetProfileData")); DNBProfileDataScanType scan_type = eProfileAll; std::string name; std::string value; @@ -4408,7 +4408,7 @@ RNBRemote::HandlePacket_SetEnableAsyncProfiling (const char *p) if (pid == INVALID_NUB_PROCESS) return SendPacket ("OK"); - StringExtractor packet(p += sizeof ("QSetEnableAsyncProfiling")); + StdStringExtractor packet(p += sizeof ("QSetEnableAsyncProfiling")); bool enable = false; uint64_t interval_usec = 0; DNBProfileDataScanType scan_type = eProfileAll; -- 2.7.4