feat: Implement string literal conversion code actions, add testing code

Author: WillLillis

src/Server.zig

@@ -1576,6 +1576,9 @@ fn codeActionHandler(server: *Server, arena: std.mem.Allocator, request: types.C
         try builder.generateCodeAction(diagnostic, &actions, &remove_capture_actions);
     }
 
+    try builder.addCodeAction(.{ .str_kind_conv = .@"string literal to multiline string" }, request, &actions);
+    try builder.addCodeAction(.{ .str_kind_conv = .@"multiline string to string literal" }, request, &actions);
+
     const Result = getRequestMetadata("textDocument/codeAction").?.Result;
     const result = try arena.alloc(std.meta.Child(std.meta.Child(Result)), actions.items.len);
     for (actions.items, result) |action, *out| {

src/features/code_actions.zig

@@ -50,6 +50,22 @@ pub const Builder = struct {
         }
     }
 
+    pub fn addCodeAction(
+        builder: *Builder,
+        kind: UserActionKind,
+        params: types.CodeActionParams,
+        actions: *std.ArrayListUnmanaged(types.CodeAction),
+    ) error{OutOfMemory}!void {
+        const loc = offsets.rangeToLoc(builder.handle.tree.source, params.range, builder.offset_encoding);
+
+        switch (kind) {
+            .str_kind_conv => |conv_kind| switch (conv_kind) {
+                .@"string literal to multiline string" => try handleStringLiteralToMultiline(builder, actions, loc),
+                .@"multiline string to string literal" => try handleMultilineStringToLiteral(builder, actions, loc),
+            },
+        }
+    }
+
     pub fn createTextEditLoc(self: *Builder, loc: offsets.Loc, new_text: []const u8) types.TextEdit {
         const range = offsets.locToRange(self.handle.tree.source, loc, self.offset_encoding);
         return types.TextEdit{ .range = range, .newText = new_text };
@@ -366,6 +382,120 @@ fn handleVariableNeverMutated(builder: *Builder, actions: *std.ArrayListUnmanage
     });
 }
 
+fn handleStringLiteralToMultiline(builder: *Builder, actions: *std.ArrayListUnmanaged(types.CodeAction), loc: offsets.Loc) !void {
+    const tokens = builder.handle.tree.tokens;
+
+    const str_tok_idx = offsets.sourceIndexToTokenIndex(builder.handle.tree, loc.start);
+    if (tokens.items(.tag)[str_tok_idx] != .string_literal) return;
+    const token_src = builder.handle.tree.tokenSlice(str_tok_idx);
+    const str_conts = token_src[1 .. token_src.len - 1]; // Omit leading and trailing '"'
+    const edit_loc_start = builder.handle.tree.tokenLocation(tokens.items(.start)[str_tok_idx], str_tok_idx).line_start;
+
+    var multiline = std.ArrayList(u8).init(builder.arena);
+    const writer = multiline.writer();
+
+    if (builder.handle.tree.tokensOnSameLine(str_tok_idx -| 1, str_tok_idx)) {
+        try writer.writeByte('\n');
+    }
+
+    var iter = std.mem.splitSequence(u8, str_conts, "\\n");
+    while (iter.next()) |line| {
+        try writer.print("\\\\{s}\n", .{line});
+    }
+
+    // remove trailing newline in cases where it's not needed
+    if (str_tok_idx + 1 < tokens.len and !builder.handle.tree.tokensOnSameLine(str_tok_idx, str_tok_idx + 1)) {
+        _ = multiline.pop();
+    }
+
+    try actions.append(builder.arena, .{
+        .title = "string literal to multiline string",
+        .kind = .@"refactor.rewrite",
+        .isPreferred = false,
+        .edit = try builder.createWorkspaceEdit(&.{
+            builder.createTextEditLoc(
+                .{
+                    .start = edit_loc_start,
+                    .end = edit_loc_start + token_src.len,
+                },
+                multiline.items,
+            ),
+        }),
+    });
+}
+
+fn handleMultilineStringToLiteral(builder: *Builder, actions: *std.ArrayListUnmanaged(types.CodeAction), loc: offsets.Loc) !void {
+    const token_tags = builder.handle.tree.tokens.items(.tag);
+    const token_starts = builder.handle.tree.tokens.items(.start);
+
+    var multiline_tok_idx = offsets.sourceIndexToTokenIndex(builder.handle.tree, loc.start);
+    if (token_tags[multiline_tok_idx] != .multiline_string_literal_line) return;
+
+    // walk up to the first multiline string literal
+    const start_tok_idx = blk: {
+        while (multiline_tok_idx > 0) : (multiline_tok_idx -= 1) {
+            if (token_tags[multiline_tok_idx] != .multiline_string_literal_line) {
+                break :blk multiline_tok_idx + 1;
+            }
+        }
+        break :blk multiline_tok_idx;
+    };
+
+    var str_literal = std.ArrayList(u8).init(builder.arena);
+    const writer = str_literal.writer();
+
+    // place string literal on same line as the left adjacent equals sign, if it's there
+    const prev_tok_idx = start_tok_idx -| 1;
+    const edit_loc_start = blk: {
+        if (token_tags[prev_tok_idx] == .equal and !builder.handle.tree.tokensOnSameLine(prev_tok_idx, start_tok_idx)) {
+            try writer.writeAll(" \"");
+            break :blk builder.handle.tree.tokenLocation(token_starts[prev_tok_idx], prev_tok_idx).line_end;
+        } else {
+            try writer.writeByte('\"');
+            break :blk builder.handle.tree.tokenLocation(token_starts[start_tok_idx], start_tok_idx).line_start;
+        }
+    };
+
+    // construct string literal out of multiline string literals
+    var curr_tok_idx = start_tok_idx;
+    var edit_loc_end: usize = undefined;
+    while (curr_tok_idx < token_tags.len and token_tags[curr_tok_idx] == .multiline_string_literal_line) : (curr_tok_idx += 1) {
+        if (curr_tok_idx > start_tok_idx) {
+            try writer.writeAll("\\n");
+        }
+        const line = builder.handle.tree.tokenSlice(curr_tok_idx);
+        const end = if (line[line.len - 1] == '\n')
+            line.len - 1
+        else
+            line.len;
+        try writer.writeAll(line[2..end]); // Omit the leading '\\', trailing '\n' (if it's there)
+        edit_loc_end = builder.handle.tree.tokenLocation(token_starts[curr_tok_idx], curr_tok_idx).line_end;
+    }
+
+    try writer.writeByte('\"');
+
+    // bring up the semicolon from the next line, if it's there
+    if (curr_tok_idx < token_tags.len and token_tags[curr_tok_idx] == .semicolon) {
+        try writer.writeByte(';');
+        edit_loc_end = builder.handle.tree.tokenLocation(token_starts[curr_tok_idx], curr_tok_idx).line_start + 1;
+    }
+
+    try actions.append(builder.arena, .{
+        .title = "multiline string to string literal",
+        .kind = .@"refactor.rewrite",
+        .isPreferred = false,
+        .edit = try builder.createWorkspaceEdit(&.{
+            builder.createTextEditLoc(
+                .{
+                    .start = edit_loc_start,
+                    .end = edit_loc_end,
+                },
+                str_literal.items,
+            ),
+        }),
+    });
+}
+
 fn detectIndentation(source: []const u8) []const u8 {
     // Essentially I'm looking for the first indentation in the file.
     var i: usize = 0;
@@ -575,6 +705,15 @@ const DiagnosticKind = union(enum) {
     }
 };
 
+const UserActionKind = union(enum) {
+    str_kind_conv: StrCat,
+
+    const StrCat = enum {
+        @"string literal to multiline string",
+        @"multiline string to string literal",
+    };
+};
+
 /// takes the location of an identifier which is part of a discard `_ = location_here;`
 /// and returns the location from '_' until ';' or null on failure
 fn getDiscardLoc(text: []const u8, loc: offsets.Loc) ?offsets.Loc {

tests/lsp_features/code_actions.zig

@@ -363,6 +363,93 @@ test "ignore autofix comment whitespace" {
     );
 }
 
+test "string literal to multiline string literal" {
+    try testUserCodeAction(
+        \\const foo = <cursor>"line one\nline two\nline three";
+    ,
+        \\const foo = 
+        \\\\line one
+        \\\\line two
+        \\\\line three
+        \\;
+    );
+    try testUserCodeAction(
+        \\const foo = "Hello, <cursor>World!\n";
+    ,
+        \\const foo = 
+        \\\\Hello, World!
+        \\\\
+        \\;
+    );
+    try testUserCodeAction(
+        \\std.debug.print(<cursor>"Hi\nHey\nHello\n", .{});
+    ,
+        \\std.debug.print(
+        \\\\Hi
+        \\\\Hey
+        \\\\Hello
+        \\\\
+        \\, .{});
+    );
+    try testUserCodeAction(
+        \\const blank = <cursor>""
+        \\;
+    ,
+        \\const blank = 
+        \\\\
+        \\;
+    );
+}
+
+test "multiline string literal to string literal" {
+    try testUserCodeAction(
+        \\const bleh =
+        \\    \\hello
+        \\    \\world<cursor>
+        \\    ++
+        \\    \\oh?
+        \\;
+    ,
+        \\const bleh = "hello\nworld"
+        \\    ++
+        \\    \\oh?
+        \\;
+    );
+    try testUserCodeAction(
+        \\std.debug.print(
+        \\\\Hi<cursor>
+        \\\\Hey
+        \\\\Hello
+        \\\\
+        \\, .{});
+    ,
+        \\std.debug.print(
+        \\"Hi\nHey\nHello\n"
+        \\, .{});
+    );
+    try testUserCodeAction(
+        \\const nums =
+        \\    \\123
+        \\    \\456<cursor>
+        \\    \\789
+        \\    ;
+    ,
+        \\const nums = "123\n456\n789";
+    );
+    try testUserCodeAction(
+        \\for (0..42) |idx| {
+        \\    std.debug.print("{}: {}\n<cursor>", .{ idx, my_foos[idx] });
+        \\}
+    ,
+        \\for (0..42) |idx| {
+        \\    std.debug.print(
+        \\\\{}: {}
+        \\\\
+        \\, .{ idx, my_foos[idx] });
+        \\}
+    );
+}
+
 fn testAutofix(before: []const u8, after: []const u8) !void {
     try testAutofixOptions(before, after, true); // diagnostics come from our AstGen fork
     try testAutofixOptions(before, after, false); // diagnostics come from calling zig ast-check
@@ -415,3 +502,51 @@ fn testAutofixOptions(before: []const u8, after: []const u8, want_zir: bool) !vo
 
     try std.testing.expectEqualStrings(after, handle.tree.source);
 }
+
+fn testUserCodeAction(source: []const u8, expected: []const u8) !void {
+    var ctx = try Context.init();
+    defer ctx.deinit();
+
+    const cursor_idx = std.mem.indexOf(u8, source, "<cursor>").?;
+    const text = try std.mem.concat(allocator, u8, &.{ source[0..cursor_idx], source[cursor_idx + "<cursor>".len ..] });
+    defer allocator.free(text);
+
+    const uri = try ctx.addDocument(text);
+    const handle = ctx.server.document_store.getHandle(uri).?;
+    const pos = offsets.indexToPosition(text, cursor_idx, ctx.server.offset_encoding);
+
+    const params = types.CodeActionParams{
+        .textDocument = .{ .uri = uri },
+        .range = .{
+            .start = pos,
+            .end = pos,
+        },
+        .context = .{ .diagnostics = &[_]zls.types.Diagnostic{} },
+    };
+
+    @setEvalBranchQuota(5000);
+    const response = try ctx.server.sendRequestSync(ctx.arena.allocator(), "textDocument/codeAction", params) orelse {
+        std.debug.print("Server returned `null` as the result\n", .{});
+        return error.InvalidResponse;
+    };
+
+    var text_edits: std.ArrayListUnmanaged(types.TextEdit) = .{};
+    defer text_edits.deinit(allocator);
+
+    for (response) |action| {
+        const code_action = action.CodeAction;
+        if (code_action.kind.? == .@"source.fixAll") continue;
+        const workspace_edit = code_action.edit.?;
+        const changes = workspace_edit.changes.?.map;
+        try std.testing.expectEqual(@as(usize, 1), changes.count());
+        try std.testing.expect(changes.contains(uri));
+
+        try text_edits.appendSlice(allocator, changes.get(uri).?);
+    }
+
+    const actual = try zls.diff.applyTextEdits(allocator, text, text_edits.items, ctx.server.offset_encoding);
+    defer allocator.free(actual);
+    try ctx.server.document_store.refreshDocument(uri, try allocator.dupeZ(u8, actual));
+
+    try std.testing.expectEqualStrings(expected, handle.tree.source);
+}