1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160
|
//===--- SourceMgrUtils.cpp - SourceMgr LSP Utils -------------------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
#include "mlir/Tools/lsp-server-support/SourceMgrUtils.h"
#include "llvm/ADT/StringExtras.h"
#include "llvm/Support/Path.h"
#include <optional>
using namespace mlir;
using namespace mlir::lsp;
//===----------------------------------------------------------------------===//
// Utils
//===----------------------------------------------------------------------===//
/// Find the end of a string whose contents start at the given `curPtr`. Returns
/// the position at the end of the string, after a terminal or invalid character
/// (e.g. `"` or `\0`).
static const char *lexLocStringTok(const char *curPtr) {
while (char c = *curPtr++) {
// Check for various terminal characters.
if (StringRef("\"\n\v\f").contains(c))
return curPtr;
// Check for escape sequences.
if (c == '\\') {
// Check a few known escapes and \xx hex digits.
if (*curPtr == '"' || *curPtr == '\\' || *curPtr == 'n' || *curPtr == 't')
++curPtr;
else if (llvm::isHexDigit(*curPtr) && llvm::isHexDigit(curPtr[1]))
curPtr += 2;
else
return curPtr;
}
}
// If we hit this point, we've reached the end of the buffer. Update the end
// pointer to not point past the buffer.
return curPtr - 1;
}
SMRange lsp::convertTokenLocToRange(SMLoc loc, StringRef identifierChars) {
if (!loc.isValid())
return SMRange();
const char *curPtr = loc.getPointer();
// Check if this is a string token.
if (*curPtr == '"') {
curPtr = lexLocStringTok(curPtr + 1);
// Otherwise, default to handling an identifier.
} else {
// Return if the given character is a valid identifier character.
auto isIdentifierChar = [=](char c) {
return isalnum(c) || c == '_' || identifierChars.contains(c);
};
while (*curPtr && isIdentifierChar(*(++curPtr)))
continue;
}
return SMRange(loc, SMLoc::getFromPointer(curPtr));
}
std::optional<std::string>
lsp::extractSourceDocComment(llvm::SourceMgr &sourceMgr, SMLoc loc) {
// This is a heuristic, and isn't intended to cover every case, but should
// cover the most common. We essentially look for a comment preceding the
// line, and if we find one, use that as the documentation.
if (!loc.isValid())
return std::nullopt;
int bufferId = sourceMgr.FindBufferContainingLoc(loc);
if (bufferId == 0)
return std::nullopt;
const char *bufferStart =
sourceMgr.getMemoryBuffer(bufferId)->getBufferStart();
StringRef buffer(bufferStart, loc.getPointer() - bufferStart);
// Pop the last line from the buffer string.
auto popLastLine = [&]() -> std::optional<StringRef> {
size_t newlineOffset = buffer.find_last_of("\n");
if (newlineOffset == StringRef::npos)
return std::nullopt;
StringRef lastLine = buffer.drop_front(newlineOffset).trim();
buffer = buffer.take_front(newlineOffset);
return lastLine;
};
// Try to pop the current line.
if (!popLastLine())
return std::nullopt;
// Try to parse a comment string from the source file.
SmallVector<StringRef> commentLines;
while (std::optional<StringRef> line = popLastLine()) {
// Check for a comment at the beginning of the line.
if (!line->startswith("//"))
break;
// Extract the document string from the comment.
commentLines.push_back(line->drop_while([](char c) { return c == '/'; }));
}
if (commentLines.empty())
return std::nullopt;
return llvm::join(llvm::reverse(commentLines), "\n");
}
bool lsp::contains(SMRange range, SMLoc loc) {
return range.Start.getPointer() <= loc.getPointer() &&
loc.getPointer() < range.End.getPointer();
}
//===----------------------------------------------------------------------===//
// SourceMgrInclude
//===----------------------------------------------------------------------===//
Hover SourceMgrInclude::buildHover() const {
Hover hover(range);
{
llvm::raw_string_ostream hoverOS(hover.contents.value);
hoverOS << "`" << llvm::sys::path::filename(uri.file()) << "`\n***\n"
<< uri.file();
}
return hover;
}
void lsp::gatherIncludeFiles(llvm::SourceMgr &sourceMgr,
SmallVectorImpl<SourceMgrInclude> &includes) {
for (unsigned i = 1, e = sourceMgr.getNumBuffers(); i < e; ++i) {
// Check to see if this file was included by the main file.
SMLoc includeLoc = sourceMgr.getBufferInfo(i + 1).IncludeLoc;
if (!includeLoc.isValid() || sourceMgr.FindBufferContainingLoc(
includeLoc) != sourceMgr.getMainFileID())
continue;
// Try to build a URI for this file path.
auto *buffer = sourceMgr.getMemoryBuffer(i + 1);
llvm::SmallString<256> path(buffer->getBufferIdentifier());
llvm::sys::path::remove_dots(path, /*remove_dot_dot=*/true);
llvm::Expected<URIForFile> includedFileURI = URIForFile::fromFile(path);
if (!includedFileURI)
continue;
// Find the end of the include token.
const char *includeStart = includeLoc.getPointer() - 2;
while (*(--includeStart) != '\"')
continue;
// Push this include.
SMRange includeRange(SMLoc::getFromPointer(includeStart), includeLoc);
includes.emplace_back(*includedFileURI, Range(sourceMgr, includeRange));
}
}
|