X-Git-Url: http://plrg.eecs.uci.edu/git/?p=folly.git;a=blobdiff_plain;f=folly%2FString-inl.h;h=f6b5ba993dffbba4d2cc318db4c269f314a0833d;hp=52b665c75e6e662d464fa65ba42dfb18999f9078;hb=acbd746f5036bf795cc62022f0114b4b08032d82;hpb=aff1262d92fbb91ce6075a5f1a5c6039ac12d8bd diff --git a/folly/String-inl.h b/folly/String-inl.h index 52b665c7..f6b5ba99 100644 --- a/folly/String-inl.h +++ b/folly/String-inl.h @@ -1,5 +1,5 @@ /* - * Copyright 2014 Facebook, Inc. + * Copyright 2017 Facebook, Inc. * * Licensed under the Apache License, Version 2.0 (the "License"); * you may not use this file except in compliance with the License. @@ -14,13 +14,14 @@ * limitations under the License. */ -#ifndef FOLLY_STRING_INL_H_ -#define FOLLY_STRING_INL_H_ +#pragma once -#include #include +#include -#ifndef FOLLY_BASE_STRING_H_ +#include + +#ifndef FOLLY_STRING_H_ #error This file may only be included from String.h #endif @@ -32,7 +33,7 @@ namespace detail { // an octal escape sequence, or 'P' if the character is printable and // should be printed as is. extern const char cEscapeTable[]; -} // namespace detail +} // namespace detail template void cEscape(StringPiece str, String& out) { @@ -51,7 +52,7 @@ void cEscape(StringPiece str, String& out) { if (e == 'P') { // printable ++p; } else if (e == 'O') { // octal - out.append(&*last, p - last); + out.append(&*last, size_t(p - last)); esc[1] = '0' + ((v >> 6) & 7); esc[2] = '0' + ((v >> 3) & 7); esc[3] = '0' + (v & 7); @@ -59,14 +60,14 @@ void cEscape(StringPiece str, String& out) { ++p; last = p; } else { // special 1-character escape - out.append(&*last, p - last); + out.append(&*last, size_t(p - last)); esc[1] = e; out.append(esc, 2); ++p; last = p; } } - out.append(&*last, p - last); + out.append(&*last, size_t(p - last)); } namespace detail { @@ -79,7 +80,7 @@ extern const char cUnescapeTable[]; // Map from the character code to the hex value, or 16 if invalid hex char. extern const unsigned char hexTable[]; -} // namespace detail +} // namespace detail template void cUnescape(StringPiece str, String& out, bool strict) { @@ -157,7 +158,7 @@ namespace detail { // 3 = space, replace with '+' in QUERY mode // 4 = percent-encode extern const unsigned char uriEscapeTable[]; -} // namespace detail +} // namespace detail template void uriEscape(StringPiece str, String& out, UriEscapeMode mode) { @@ -178,12 +179,12 @@ void uriEscape(StringPiece str, String& out, UriEscapeMode mode) { if (LIKELY(discriminator <= minEncode)) { ++p; } else if (mode == UriEscapeMode::QUERY && discriminator == 3) { - out.append(&*last, p - last); + out.append(&*last, size_t(p - last)); out.push_back('+'); ++p; last = p; } else { - out.append(&*last, p - last); + out.append(&*last, size_t(p - last)); esc[1] = hexValues[v >> 4]; esc[2] = hexValues[v & 0x0f]; out.append(esc, 3); @@ -191,7 +192,7 @@ void uriEscape(StringPiece str, String& out, UriEscapeMode mode) { last = p; } } - out.append(&*last, p - last); + out.append(&*last, size_t(p - last)); } template @@ -203,7 +204,6 @@ void uriUnescape(StringPiece str, String& out, UriEscapeMode mode) { // this is faster than calling push_back repeatedly. while (p != str.end()) { char c = *p; - unsigned char v = static_cast(v); switch (c) { case '%': { @@ -215,7 +215,7 @@ void uriUnescape(StringPiece str, String& out, UriEscapeMode mode) { if (UNLIKELY(h1 == 16 || h2 == 16)) { throw std::invalid_argument("invalid percent encode sequence"); } - out.append(&*last, p - last); + out.append(&*last, size_t(p - last)); out.push_back((h1 << 4) | h2); p += 3; last = p; @@ -223,19 +223,20 @@ void uriUnescape(StringPiece str, String& out, UriEscapeMode mode) { } case '+': if (mode == UriEscapeMode::QUERY) { - out.append(&*last, p - last); + out.append(&*last, size_t(p - last)); out.push_back(' '); ++p; last = p; break; } // else fallthrough + FOLLY_FALLTHROUGH; default: ++p; break; } } - out.append(&*last, p - last); + out.append(&*last, size_t(p - last)); } namespace detail { @@ -265,29 +266,6 @@ inline char delimFront(StringPiece s) { return *s.start(); } -/* - * These output conversion templates allow us to support multiple - * output string types, even when we are using an arbitrary - * OutputIterator. - */ -template struct OutputConverter {}; - -template<> struct OutputConverter { - std::string operator()(StringPiece sp) const { - return sp.toString(); - } -}; - -template<> struct OutputConverter { - fbstring operator()(StringPiece sp) const { - return sp.toFbstring(); - } -}; - -template<> struct OutputConverter { - StringPiece operator()(StringPiece sp) const { return sp; } -}; - /* * Shared implementation for all the split() overloads. * @@ -297,7 +275,7 @@ template<> struct OutputConverter { * * @param ignoreEmpty iff true, don't copy empty segments to output */ -template +template void internalSplit(DelimT delim, StringPiece sp, OutputIterator out, bool ignoreEmpty) { assert(sp.empty() || sp.start() != nullptr); @@ -306,11 +284,9 @@ void internalSplit(DelimT delim, StringPiece sp, OutputIterator out, const size_t strSize = sp.size(); const size_t dSize = delimSize(delim); - OutputConverter conv; - if (dSize > strSize || dSize == 0) { if (!ignoreEmpty || strSize > 0) { - *out++ = conv(sp); + *out++ = to(sp); } return; } @@ -320,12 +296,12 @@ void internalSplit(DelimT delim, StringPiece sp, OutputIterator out, ignoreEmpty); } - int tokenStartPos = 0; - int tokenSize = 0; - for (int i = 0; i <= strSize - dSize; ++i) { + size_t tokenStartPos = 0; + size_t tokenSize = 0; + for (size_t i = 0; i <= strSize - dSize; ++i) { if (atDelim(&s[i], delim)) { if (!ignoreEmpty || tokenSize > 0) { - *out++ = conv(StringPiece(&s[tokenStartPos], tokenSize)); + *out++ = to(sp.subpiece(tokenStartPos, tokenSize)); } tokenStartPos = i + dSize; @@ -335,48 +311,36 @@ void internalSplit(DelimT delim, StringPiece sp, OutputIterator out, ++tokenSize; } } - + tokenSize = strSize - tokenStartPos; if (!ignoreEmpty || tokenSize > 0) { - tokenSize = strSize - tokenStartPos; - *out++ = conv(StringPiece(&s[tokenStartPos], tokenSize)); + *out++ = to(sp.subpiece(tokenStartPos, tokenSize)); } } -template StringPiece prepareDelim(const String& s) { +template StringPiece prepareDelim(const String& s) { return StringPiece(s); } inline char prepareDelim(char c) { return c; } -template -struct convertTo { - template - static Dst from(const Src& src) { return folly::to(src); } - static Dst from(const Dst& src) { return src; } -}; - -template -typename std::enable_if::value, bool>::type -splitFixed(const Delim& delimiter, - StringPiece input, - OutputType& out) { +template +bool splitFixed(const Delim& delimiter, StringPiece input, OutputType& output) { + static_assert( + exact || std::is_same::value || + IsSomeString::value, + "split() requires that the last argument be a string type"); if (exact && UNLIKELY(std::string::npos != input.find(delimiter))) { return false; } - out = convertTo::from(input); + output = folly::to(input); return true; } -template -typename std::enable_if::value, bool>::type -splitFixed(const Delim& delimiter, - StringPiece input, - OutputType& outHead, - OutputTypes&... outTail) { +template +bool splitFixed( + const Delim& delimiter, + StringPiece input, + OutputType& outHead, + OutputTypes&... outTail) { size_t cut = input.find(delimiter); if (UNLIKELY(cut == std::string::npos)) { return false; @@ -385,17 +349,17 @@ splitFixed(const Delim& delimiter, StringPiece tail(input.begin() + cut + detail::delimSize(delimiter), input.end()); if (LIKELY(splitFixed(delimiter, tail, outTail...))) { - outHead = convertTo::from(head); + outHead = folly::to(head); return true; } return false; } -} +} // namespace detail ////////////////////////////////////////////////////////////////////// -template +template void split(const Delim& delimiter, const String& input, std::vector& out, @@ -407,7 +371,7 @@ void split(const Delim& delimiter, ignoreEmpty); } -template +template void split(const Delim& delimiter, const String& input, fbvector& out, @@ -419,8 +383,11 @@ void split(const Delim& delimiter, ignoreEmpty); } -template +template < + class OutputValueType, + class Delim, + class String, + class OutputIterator> void splitTo(const Delim& delimiter, const String& input, OutputIterator out, @@ -432,20 +399,13 @@ void splitTo(const Delim& delimiter, ignoreEmpty); } -template -typename std::enable_if::value, bool>::type -split(const Delim& delimiter, - StringPiece input, - OutputType& outHead, - OutputTypes&... outTail) { +template +typename std::enable_if< + AllConvertible::value && sizeof...(OutputTypes) >= 1, + bool>::type +split(const Delim& delimiter, StringPiece input, OutputTypes&... outputs) { return detail::splitFixed( - detail::prepareDelim(delimiter), - input, - outHead, - outTail...); + detail::prepareDelim(delimiter), input, outputs...); } namespace detail { @@ -517,7 +477,7 @@ internalJoin(Delim delimiter, internalJoinAppend(delimiter, begin, end, output); } -} // namespace detail +} // namespace detail template void join(const Delim& delimiter, @@ -531,8 +491,11 @@ void join(const Delim& delimiter, output); } -template -void backslashify(const String1& input, String2& output, bool hex_style) { +template +void backslashify( + folly::StringPiece input, + OutputString& output, + bool hex_style) { static const char hexValues[] = "0123456789abcdef"; output.clear(); output.reserve(3 * input.size()); @@ -544,14 +507,21 @@ void backslashify(const String1& input, String2& output, bool hex_style) { if (hex_style) { hex_append = true; } else { - if (c == '\r') output += 'r'; - else if (c == '\n') output += 'n'; - else if (c == '\t') output += 't'; - else if (c == '\a') output += 'a'; - else if (c == '\b') output += 'b'; - else if (c == '\0') output += '0'; - else if (c == '\\') output += '\\'; - else { + if (c == '\r') { + output += 'r'; + } else if (c == '\n') { + output += 'n'; + } else if (c == '\t') { + output += 't'; + } else if (c == '\a') { + output += 'a'; + } else if (c == '\b') { + output += 'b'; + } else if (c == '\0') { + output += '0'; + } else if (c == '\\') { + output += '\\'; + } else { hex_append = true; } } @@ -568,8 +538,8 @@ void backslashify(const String1& input, String2& output, bool hex_style) { template void humanify(const String1& input, String2& output) { - int numUnprintable = 0; - int numPrintablePrefix = 0; + size_t numUnprintable = 0; + size_t numPrintablePrefix = 0; for (unsigned char c : input) { if (c < 0x20 || c > 0x7e || c == '\\') { ++numUnprintable; @@ -610,15 +580,17 @@ void humanify(const String1& input, String2& output) { } } -template +template bool hexlify(const InputString& input, OutputString& output, bool append_output) { - if (!append_output) output.clear(); + if (!append_output) { + output.clear(); + } static char hexValues[] = "0123456789abcdef"; - int j = output.size(); + auto j = output.size(); output.resize(2 * input.size() + output.size()); - for (int i = 0; i < input.size(); ++i) { + for (size_t i = 0; i < input.size(); ++i) { int ch = input[i]; output[j++] = hexValues[(ch >> 4) & 0xf]; output[j++] = hexValues[ch & 0xf]; @@ -626,24 +598,19 @@ bool hexlify(const InputString& input, OutputString& output, return true; } -template +template bool unhexlify(const InputString& input, OutputString& output) { if (input.size() % 2 != 0) { return false; } output.resize(input.size() / 2); int j = 0; - auto unhex = [](char c) -> int { - return c >= '0' && c <= '9' ? c - '0' : - c >= 'A' && c <= 'F' ? c - 'A' + 10 : - c >= 'a' && c <= 'f' ? c - 'a' + 10 : - -1; - }; - - for (int i = 0; i < input.size(); i += 2) { - int highBits = unhex(input[i]); - int lowBits = unhex(input[i + 1]); - if (highBits < 0 || lowBits < 0) { + + for (size_t i = 0; i < input.size(); i += 2) { + int highBits = detail::hexTable[static_cast(input[i])]; + int lowBits = detail::hexTable[static_cast(input[i + 1])]; + if ((highBits | lowBits) & 0x10) { + // One of the characters wasn't a hex digit return false; } output[j++] = (highBits << 4) + lowBits; @@ -658,7 +625,7 @@ namespace detail { */ size_t hexDumpLine(const void* ptr, size_t offset, size_t size, std::string& line); -} // namespace detail +} // namespace detail template void hexDump(const void* ptr, size_t size, OutIt out) { @@ -670,6 +637,4 @@ void hexDump(const void* ptr, size_t size, OutIt out) { } } -} // namespace folly - -#endif /* FOLLY_STRING_INL_H_ */ +} // namespace folly