microsoft/TypeScript
1
0
Fork 0
mirror of https://github.com/microsoft/TypeScript.git synced 2026-05-15 04:00:52 -05:00
Code Issues Packages Projects Releases 212 Wiki Activity

Provide better error recovery when we encounter merge markers in the source.

Browse Source 
Previously we would just treat each merge marker as trivia and then continue
scanning and parsing like normal.  This worked well in some scenarios, but
fell down in others like:

```
class C {
    public foo() {
<<<<<<< HEAD
        this.bar();
    }
=======
        this.baz();
    }
>>>>>>> Branch

    public bar() { }
}
```

The problem stems from the previous approach trying to incorporate both branches of the merge into
the final tree.  In a case like this, that approach breaks down entirely.  The the parser ends up
seeing the close curly in both included sections, and it considers the class finished.  Then, it
starts erroring when it encounters "public bar()".

The fix is to only incorporate one of these sections into the tree.  Specifically, we only include
the first section.  The second sectoin is treated like trivia and does not affect the parse at all.
To make the experience more pleasant we do *lexically* classify the second section.  That way it
does not appear as just plain black text in the editor.  Instead, it will have appropriate lexicla
classifications for keywords, literals, comments, operators, punctuation, etc.  However, any syntactic
or semantic feature will not work in the second block due to this being trivia as far as any feature
is concerned.

This experience is still much better than what we had originally (where merge markers would absolutely)
destroy the parse tree.  And it is better than what we checked in last week, which could easily create
a borked tree for many types of merges.

Now, almost all merges should still leave the tree in good shape.  All LS features will work in the
first section, and lexical classification will work in the second.
This commit is contained in:
Cyrus Najmabadi
2014-12-18 19:18:13 -08:00
parent 828b33aae7
commit 48bef4698b
7 changed files with 270 additions and 163 deletions
Download Patch File Download Diff File Expand all files Collapse all files

50
src/compiler/scanner.ts
View File

@@ -7,10 +7,6 @@ module ts {
(message: DiagnosticMessage, length: number): void;
}
export interface CommentCallback {
(pos: number, end: number): void;
}
export interface Scanner {
getStartPos(): number;
getToken(): SyntaxKind;
@@ -396,8 +392,10 @@ module ts {
var mergeConflictMarkerLength = "<<<<<<<".length;
function isConflictMarkerTrivia(text: string, pos: number) {
Debug.assert(pos >= 0);
// Conflict markers must be at the start of a line.
if (pos > 0 && isLineBreak(text.charCodeAt(pos - 1))) {
if (pos === 0 || isLineBreak(text.charCodeAt(pos - 1))) {
var ch = text.charCodeAt(pos);
if ((pos + mergeConflictMarkerLength) < text.length) {
@@ -415,10 +413,31 @@ module ts {
return false;
}
function scanConflictMarkerTrivia(text: string, pos: number) {
var len = text.length;
while (pos < len && !isLineBreak(text.charCodeAt(pos))) {
pos++;
function scanConflictMarkerTrivia(text: string, pos: number, error?: ErrorCallback) {
if (error) {
error(Diagnostics.Merge_conflict_marker_encountered, mergeConflictMarkerLength);
}
var ch = text.charCodeAt(pos);
if (ch === CharacterCodes.lessThan || ch === CharacterCodes.greaterThan) {
var len = text.length;
while (pos < len && !isLineBreak(text.charCodeAt(pos))) {
pos++;
}
}
else {
Debug.assert(ch === CharacterCodes.equals);
// Consume everything from the start of the mid-conlict marker to the start of the next
// end-conflict marker.
var len = text.length;
while (pos < len) {
var ch = text.charCodeAt(pos);
if (ch === CharacterCodes.greaterThan && isConflictMarkerTrivia(text, pos)) {
break;
}
pos++;
}
}
return pos;
@@ -1057,8 +1076,7 @@ module ts {
return pos++, token = SyntaxKind.SemicolonToken;
case CharacterCodes.lessThan:
if (isConflictMarkerTrivia(text, pos)) {
mergeConflictError();
pos = scanConflictMarkerTrivia(text, pos);
pos = scanConflictMarkerTrivia(text, pos, error);
if (skipTrivia) {
continue;
}
@@ -1079,8 +1097,7 @@ module ts {
return pos++, token = SyntaxKind.LessThanToken;
case CharacterCodes.equals:
if (isConflictMarkerTrivia(text, pos)) {
mergeConflictError();
pos = scanConflictMarkerTrivia(text, pos);
pos = scanConflictMarkerTrivia(text, pos, error);
if (skipTrivia) {
continue;
}
@@ -1101,8 +1118,7 @@ module ts {
return pos++, token = SyntaxKind.EqualsToken;
case CharacterCodes.greaterThan:
if (isConflictMarkerTrivia(text, pos)) {
mergeConflictError();
pos = scanConflictMarkerTrivia(text, pos);
pos = scanConflictMarkerTrivia(text, pos, error);
if (skipTrivia) {
continue;
}
@@ -1171,10 +1187,6 @@ module ts {
}
}
function mergeConflictError() {
error(Diagnostics.Merge_conflict_marker_encountered, mergeConflictMarkerLength);
}
function reScanGreaterToken(): SyntaxKind {
if (token === SyntaxKind.GreaterThanToken) {
if (text.charCodeAt(pos) === CharacterCodes.greaterThan) {

224
src/services/services.ts
View File

@@ -2155,7 +2155,7 @@ module ts {
// invalid identifier name. We need to check if whatever was inside the quotes is actually a valid identifier name.
displayName = displayName.substring(1, displayName.length - 1);
}
var isValid = isIdentifierStart(displayName.charCodeAt(0), target);
for (var i = 1, n = displayName.length; isValid && i < n; i++) {
isValid = isIdentifierPart(displayName.charCodeAt(i), target);
@@ -2206,7 +2206,7 @@ module ts {
// Completion not allowed inside comments, bail out if this is the case
var insideComment = isInsideComment(sourceFile, currentToken, position);
host.log("getCompletionsAtPosition: Is inside comment: " + (new Date().getTime() - start));
if (insideComment) {
host.log("Returning an empty list because completion was inside a comment.");
return undefined;
@@ -2593,7 +2593,7 @@ module ts {
if (flags & SymbolFlags.TypeAlias) return ScriptElementKind.typeElement;
if (flags & SymbolFlags.Interface) return ScriptElementKind.interfaceElement;
if (flags & SymbolFlags.TypeParameter) return ScriptElementKind.typeParameterElement;
var result = getSymbolKindOfConstructorPropertyMethodAccessorFunctionOrVar(symbol, flags, typeResolver, location);
if (result === ScriptElementKind.unknown) {
if (flags & SymbolFlags.TypeParameter) return ScriptElementKind.typeParameterElement;
@@ -2656,7 +2656,7 @@ module ts {
return ScriptElementKind.unknown;
}
function getTypeKind(type: Type): string {
var flags = type.getFlags();
@@ -2730,7 +2730,7 @@ module ts {
if (location.parent && location.parent.kind === SyntaxKind.PropertyAccessExpression) {
var right = (<PropertyAccessExpression>location.parent).name;
// Either the location is on the right of a property access, or on the left and the right is missing
if (right === location || (right && right.getFullWidth() === 0)){
if (right === location || (right && right.getFullWidth() === 0)) {
location = location.parent;
}
}
@@ -2974,7 +2974,7 @@ module ts {
symbolFlags & SymbolFlags.Method ||
symbolFlags & SymbolFlags.Constructor ||
symbolFlags & SymbolFlags.Signature ||
symbolFlags & SymbolFlags.Accessor ||
symbolFlags & SymbolFlags.Accessor ||
symbolKind === ScriptElementKind.memberFunctionElement) {
var allSignatures = type.getCallSignatures();
addSignatureDisplayParts(allSignatures[0], allSignatures);
@@ -3333,7 +3333,7 @@ module ts {
if (!hasKind(ifStatement.elseStatement, SyntaxKind.IfStatement)) {
break
}
ifStatement = <IfStatement>ifStatement.elseStatement;
}
@@ -3355,7 +3355,7 @@ module ts {
break;
}
}
if (shouldHighlightNextKeyword) {
result.push({
fileName: filename,
@@ -3394,7 +3394,7 @@ module ts {
return map(keywords, getReferenceEntryFromNode);
}
function getThrowOccurrences(throwStatement: ThrowStatement) {
var owner = getThrowStatementOwner(throwStatement);
@@ -3403,7 +3403,7 @@ module ts {
}
var keywords: Node[] = [];
forEach(aggregateOwnedThrowStatements(owner), throwStatement => {
pushKeywordIf(keywords, throwStatement.getFirstToken(), SyntaxKind.ThrowKeyword);
});
@@ -3415,7 +3415,7 @@ module ts {
pushKeywordIf(keywords, returnStatement.getFirstToken(), SyntaxKind.ReturnKeyword);
});
}
return map(keywords, getReferenceEntryFromNode);
}
@@ -3551,7 +3551,7 @@ module ts {
return map(keywords, getReferenceEntryFromNode);
}
function getBreakOrContinueStatementOccurences(breakOrContinueStatement: BreakOrContinueStatement): ReferenceEntry[]{
function getBreakOrContinueStatementOccurences(breakOrContinueStatement: BreakOrContinueStatement): ReferenceEntry[] {
var owner = getBreakOrContinueOwner(breakOrContinueStatement);
if (owner) {
@@ -3599,7 +3599,7 @@ module ts {
if (statement.kind === SyntaxKind.ContinueStatement) {
continue;
}
// Fall through.
// Fall through.
case SyntaxKind.ForStatement:
case SyntaxKind.ForInStatement:
case SyntaxKind.WhileStatement:
@@ -4024,13 +4024,13 @@ module ts {
* searchLocation: a node where the search value
*/
function getReferencesInNode(container: Node,
searchSymbol: Symbol,
searchText: string,
searchLocation: Node,
searchMeaning: SemanticMeaning,
findInStrings: boolean,
findInComments: boolean,
result: ReferenceEntry[]): void {
searchSymbol: Symbol,
searchText: string,
searchLocation: Node,
searchMeaning: SemanticMeaning,
findInStrings: boolean,
findInComments: boolean,
result: ReferenceEntry[]): void {
var sourceFile = container.getSourceFile();
var tripleSlashDirectivePrefixRegex = /^\/\/\/\s*</
@@ -4110,7 +4110,7 @@ module ts {
}
}
function getReferencesForSuperKeyword(superKeyword: Node): ReferenceEntry[]{
function getReferencesForSuperKeyword(superKeyword: Node): ReferenceEntry[] {
var searchSpaceNode = getSuperContainer(superKeyword);
if (!searchSpaceNode) {
return undefined;
@@ -4184,7 +4184,7 @@ module ts {
if (isExternalModule(<SourceFile>searchSpaceNode)) {
return undefined;
}
// Fall through
// Fall through
case SyntaxKind.FunctionDeclaration:
case SyntaxKind.FunctionExpression:
break;
@@ -4861,26 +4861,105 @@ module ts {
fileName = normalizeSlashes(fileName);
var sourceFile = getCurrentSourceFile(fileName);
// Make a scanner we can get trivia from.
var triviaScanner = createScanner(ScriptTarget.Latest, /*skipTrivia:*/ false, sourceFile.text);
var mergeConflictScanner = createScanner(ScriptTarget.Latest, /*skipTrivia:*/ false, sourceFile.text);
var result: ClassifiedSpan[] = [];
processElement(sourceFile);
return result;
function classifyComment(comment: CommentRange) {
var width = comment.end - comment.pos;
if (textSpanIntersectsWith(span, comment.pos, width)) {
function classifyLeadingTrivia(token: Node): void {
var tokenStart = skipTrivia(sourceFile.text, token.pos, /*stopAfterLineBreak:*/ false);
if (tokenStart === token.pos) {
return;
}
// token has trivia. Classify them appropriately.
triviaScanner.setTextPos(token.pos);
while (true) {
var start = triviaScanner.getTextPos();
var kind = triviaScanner.scan();
var end = triviaScanner.getTextPos();
var width = end - start;
if (textSpanIntersectsWith(span, start, width)) {
if (!isTrivia(kind)) {
return;
}
if (isComment(kind)) {
// Simple comment. Just add as is.
result.push({
textSpan: createTextSpan(start, width),
classificationType: ClassificationTypeNames.comment
})
continue;
}
if (kind === SyntaxKind.ConflictMarkerTrivia) {
var text = sourceFile.text;
var ch = text.charCodeAt(start);
// for the <<<<<<< and >>>>>>> markers, we just add them as in as
// comments in the classification stream.
if (ch === CharacterCodes.lessThan || ch === CharacterCodes.greaterThan) {
result.push({
textSpan: createTextSpan(start, width),
classificationType: ClassificationTypeNames.comment
});
continue;
}
// for the ======== add a comment for the first line, and then lex all
// subsequent lines up until the end of the conflict marker.
Debug.assert(ch === CharacterCodes.equals);
classifyDisabledCode(text, start, end);
}
}
}
}
function classifyDisabledCode(text: string, start: number, end: number) {
// Classify the line that the ======= marker is on as a comment. Then just lex
// all further tokens and add them to the result.
for (var i = start; i < end; i++) {
if (isLineBreak(text.charCodeAt(i))) {
break;
}
}
result.push({
textSpan: createTextSpanFromBounds(start, i),
classificationType: ClassificationTypeNames.comment
});
mergeConflictScanner.setTextPos(i);
while (mergeConflictScanner.getTextPos() < end) {
classifyDisabledCodeToken();
}
}
function classifyDisabledCodeToken() {
var start = mergeConflictScanner.getTextPos();
var tokenKind = mergeConflictScanner.scan();
var end = mergeConflictScanner.getTextPos();
var type = classifyTokenType(tokenKind);
if (type) {
result.push({
textSpan: createTextSpan(comment.pos, width),
classificationType: ClassificationTypeNames.comment
textSpan: createTextSpanFromBounds(start, end),
classificationType: type
});
}
}
function classifyToken(token: Node): void {
forEach(getLeadingCommentRanges(sourceFile.text, token.getFullStart()), classifyComment);
classifyLeadingTrivia(token);
if (token.getWidth() > 0) {
var type = classifyTokenType(token);
var type = classifyTokenType(token.kind, token);
if (type) {
result.push({
textSpan: createTextSpan(token.getStart(), token.getWidth()),
@@ -4888,12 +4967,9 @@ module ts {
});
}
}
forEach(getTrailingCommentRanges(sourceFile.text, token.getEnd()), classifyComment);
}
function classifyTokenType(token: Node): string {
var tokenKind = token.kind;
function classifyTokenType(tokenKind: SyntaxKind, token?: Node): string {
if (isKeyword(tokenKind)) {
return ClassificationTypeNames.keyword;
}
@@ -4903,23 +4979,24 @@ module ts {
if (tokenKind === SyntaxKind.LessThanToken || tokenKind === SyntaxKind.GreaterThanToken) {
// If the node owning the token has a type argument list or type parameter list, then
// we can effectively assume that a '<' and '>' belong to those lists.
if (getTypeArgumentOrTypeParameterList(token.parent)) {
if (token && getTypeArgumentOrTypeParameterList(token.parent)) {
return ClassificationTypeNames.punctuation;
}
}
if (isPunctuation(token.kind)) {
if (isPunctuation(tokenKind)) {
// the '=' in a variable declaration is special cased here.
if (token.parent.kind === SyntaxKind.BinaryExpression ||
token.parent.kind === SyntaxKind.VariableDeclaration ||
token.parent.kind === SyntaxKind.PrefixUnaryExpression ||
token.parent.kind === SyntaxKind.PostfixUnaryExpression ||
token.parent.kind === SyntaxKind.ConditionalExpression) {
return ClassificationTypeNames.operator;
}
else {
return ClassificationTypeNames.punctuation;
if (token) {
if (token.parent.kind === SyntaxKind.BinaryExpression ||
token.parent.kind === SyntaxKind.VariableDeclaration ||
token.parent.kind === SyntaxKind.PrefixUnaryExpression ||
token.parent.kind === SyntaxKind.PostfixUnaryExpression ||
token.parent.kind === SyntaxKind.ConditionalExpression) {
return ClassificationTypeNames.operator;
}
}
return ClassificationTypeNames.punctuation;
}
else if (tokenKind === SyntaxKind.NumericLiteral) {
return ClassificationTypeNames.numericLiteral;
@@ -4936,35 +5013,37 @@ module ts {
return ClassificationTypeNames.stringLiteral;
}
else if (tokenKind === SyntaxKind.Identifier) {
switch (token.parent.kind) {
case SyntaxKind.ClassDeclaration:
if ((<ClassDeclaration>token.parent).name === token) {
return ClassificationTypeNames.className;
}
return;
case SyntaxKind.TypeParameter:
if ((<TypeParameterDeclaration>token.parent).name === token) {
return ClassificationTypeNames.typeParameterName;
}
return;
case SyntaxKind.InterfaceDeclaration:
if ((<InterfaceDeclaration>token.parent).name === token) {
return ClassificationTypeNames.interfaceName;
}
return;
case SyntaxKind.EnumDeclaration:
if ((<EnumDeclaration>token.parent).name === token) {
return ClassificationTypeNames.enumName;
}
return;
case SyntaxKind.ModuleDeclaration:
if ((<ModuleDeclaration>token.parent).name === token) {
return ClassificationTypeNames.moduleName;
}
return;
default:
return ClassificationTypeNames.text;
if (token) {
switch (token.parent.kind) {
case SyntaxKind.ClassDeclaration:
if ((<ClassDeclaration>token.parent).name === token) {
return ClassificationTypeNames.className;
}
return;
case SyntaxKind.TypeParameter:
if ((<TypeParameterDeclaration>token.parent).name === token) {
return ClassificationTypeNames.typeParameterName;
}
return;
case SyntaxKind.InterfaceDeclaration:
if ((<InterfaceDeclaration>token.parent).name === token) {
return ClassificationTypeNames.interfaceName;
}
return;
case SyntaxKind.EnumDeclaration:
if ((<EnumDeclaration>token.parent).name === token) {
return ClassificationTypeNames.enumName;
}
return;
case SyntaxKind.ModuleDeclaration:
if ((<ModuleDeclaration>token.parent).name === token) {
return ClassificationTypeNames.moduleName;
}
return;
}
}
return ClassificationTypeNames.text;
}
}
@@ -5493,7 +5572,6 @@ module ts {
var start = scanner.getTokenPos();
var end = scanner.getTextPos();
// add the token
addResult(end - start, classFromKind(token));
if (end >= text.length) {