terminal: OSC parser

src/terminal/Parser.zig

@@ -7,6 +7,7 @@ const Parser = @This();
 const std = @import("std");
 const testing = std.testing;
 const table = @import("parse_table.zig").table;
+const osc = @import("osc.zig");
 
 const log = std.log.scoped(.parser);
 
@@ -66,6 +67,9 @@ pub const Action = union(enum) {
     /// Execute the ESC command.
     esc_dispatch: ESC,
 
+    /// Execute the OSC command.
+    osc_dispatch: osc.Command,
+
     pub const CSI = struct {
         intermediates: []u8,
         params: []u16,
@@ -95,6 +99,9 @@ params_idx: u8 = 0,
 param_acc: u16 = 0,
 param_acc_idx: u8 = 0,
 
+/// Parser for OSC sequences
+osc_parser: osc.Parser = .{},
+
 pub fn init() Parser {
     return .{};
 }
@@ -126,20 +133,28 @@ pub fn next(self: *Parser, c: u8) [3]?Action {
     // 2. transition action
     // 3. entry action to new state
     return [3]?Action{
-        switch (self.state) {
-            .osc_string => @panic("TODO"), // TODO: osc_end
+        // Exit depends on current state
+        if (self.state == next_state) null else switch (self.state) {
+            .osc_string => if (self.osc_parser.end()) |cmd|
+                Action{ .osc_dispatch = cmd }
+            else
+                null,
             .dcs_passthrough => @panic("TODO"), // TODO: unhook
             else => null,
         },
 
         self.doAction(action, c),
 
-        switch (next_state) {
+        // Entry depends on new state
+        if (self.state == next_state) null else switch (next_state) {
             .escape, .dcs_entry, .csi_entry => clear: {
                 self.clear();
                 break :clear null;
             },
-            .osc_string => @panic("TODO"), // TODO: osc_start
+            .osc_string => osc_string: {
+                self.osc_parser.reset();
+                break :osc_string null;
+            },
             .dcs_passthrough => @panic("TODO"), // TODO: hook
             else => null,
         },
@@ -147,7 +162,6 @@ pub fn next(self: *Parser, c: u8) [3]?Action {
 }
 
 fn doAction(self: *Parser, action: TransitionAction, c: u8) ?Action {
-    _ = self;
     return switch (action) {
         .none, .ignore => null,
         .print => Action{ .print = c },
@@ -191,6 +205,10 @@ fn doAction(self: *Parser, action: TransitionAction, c: u8) ?Action {
             // The client is expected to perform no action.
             break :param null;
         },
+        .osc_put => osc_put: {
+            self.osc_parser.next(c);
+            break :osc_put null;
+        },
         .csi_dispatch => csi_dispatch: {
             // Finalize parameters if we have one
             if (self.param_acc_idx > 0) {
@@ -309,3 +327,25 @@ test "csi: ESC [ 1 ; 4 H" {
         try testing.expectEqual(@as(u16, 4), d.params[1]);
     }
 }
+
+test "osc: change window title" {
+    var p = init();
+    _ = p.next(0x1B);
+    _ = p.next(']');
+    _ = p.next('0');
+    _ = p.next(';');
+    _ = p.next('a');
+    _ = p.next('b');
+    _ = p.next('c');
+
+    {
+        const a = p.next(0x07); // BEL
+        try testing.expect(p.state == .ground);
+        try testing.expect(a[0].? == .osc_dispatch);
+        try testing.expect(a[1] == null);
+        try testing.expect(a[2] == null);
+
+        const cmd = a[0].?.osc_dispatch;
+        try testing.expect(cmd == .change_window_title);
+    }
+}

src/terminal/Terminal.zig

@@ -137,6 +137,7 @@ pub fn appendChar(self: *Terminal, alloc: Allocator, c: u8) !void {
             .execute => |code| try self.execute(alloc, code),
             .csi_dispatch => |csi| try self.csiDispatch(alloc, csi),
             .esc_dispatch => |esc| try self.escDispatch(alloc, esc),
+            .osc_dispatch => |cmd| log.warn("unhandled OSC: {}", .{cmd}),
         }
     }
 }
@@ -626,6 +627,7 @@ fn getOrPutCell(self: *Terminal, alloc: Allocator, x: usize, y: usize) !*Cell {
 }
 
 test {
+    _ = @import("osc.zig");
     _ = Parser;
     _ = Tabstops;
 }

src/terminal/osc.zig

@@ -0,0 +1,119 @@
+//! OSC (Operating System Command) related functions and types. OSC is
+//! another set of control sequences for terminal programs that start with
+//! "ESC ]". Unlike CSI or standard ESC sequences, they may contain strings
+//! and other irregular formatting so a dedicated parser is created to handle it.
+const osc = @This();
+
+const std = @import("std");
+
+const log = std.log.scoped(.osc);
+
+pub const Command = union(enum) {
+    /// Set the window title of the terminal
+    ///
+    /// If title mode 0  is set text is expect to be hex encoded (i.e. utf-8
+    /// with each code unit further encoded with two hex digets).
+    ///
+    /// If title mode 2 is set or the terminal is setup for unconditional
+    /// utf-8 titles text is interpreted as utf-8. Else text is interpreted
+    /// as latin1.
+    change_window_title: []const u8,
+};
+
+pub const Parser = struct {
+    state: State = .empty,
+    command: Command = undefined,
+    param_str: ?*[]const u8 = null,
+    buf: [MAX_BUF]u8 = undefined,
+    buf_start: usize = 0,
+    buf_idx: usize = 0,
+    complete: bool = false,
+
+    // Maximum length of a single OSC command. This is the full OSC command
+    // sequence length (excluding ESC ]). This is arbitrary, I couldn't find
+    // any definitive resource on how long this should be.
+    const MAX_BUF = 2048;
+
+    pub const State = enum {
+        empty,
+        invalid,
+        @"0",
+        string,
+    };
+
+    /// Reset the parser start.
+    pub fn reset(self: *Parser) void {
+        self.state = .empty;
+        self.param_str = null;
+        self.buf_start = 0;
+        self.buf_idx = 0;
+        self.complete = false;
+    }
+
+    /// Consume the next character c and advance the parser state.
+    pub fn next(self: *Parser, c: u8) void {
+        // We store everything in the buffer so we can do a better job
+        // logging if we get to an invalid command.
+        self.buf[self.buf_idx] = c;
+        self.buf_idx += 1;
+
+        log.info("state = {} c = {x}", .{ self.state, c });
+
+        switch (self.state) {
+            // Ignore, we're in some invalid state and we can't possibly
+            // do anything reasonable.
+            .invalid => {},
+
+            .empty => switch (c) {
+                '0' => self.state = .@"0",
+                else => self.state = .invalid,
+            },
+
+            .@"0" => switch (c) {
+                ';' => {
+                    self.command = .{ .change_window_title = undefined };
+
+                    self.state = .string;
+                    self.param_str = &self.command.change_window_title;
+                    self.buf_start = self.buf_idx;
+                },
+                else => self.state = .invalid,
+            },
+
+            .string => {
+                // Complete once we receive one character since we have
+                // at least SOME value for the expected string value.
+                self.complete = true;
+            },
+        }
+    }
+
+    /// End the sequence and return the command, if any. If the return value
+    /// is null, then no valid command was found.
+    pub fn end(self: Parser) ?Command {
+        if (!self.complete) {
+            log.warn("invalid OSC command: {s}", .{self.buf[0..self.buf_idx]});
+            return null;
+        }
+
+        // If we have an expected string parameter, fill it in.
+        if (self.param_str) |param_str| {
+            param_str.* = self.buf[self.buf_start..self.buf_idx];
+        }
+
+        return self.command;
+    }
+};
+
+test "OSC: change_window_title" {
+    const testing = std.testing;
+
+    var p: Parser = .{};
+    p.next('0');
+    p.next(';');
+    p.next('a');
+    p.next('b');
+    const cmd = p.end().?;
+    try testing.expect(cmd == .change_window_title);
+    try testing.expectEqualStrings("ab", cmd.change_window_title);
+}

src/terminal/parse_table.zig

@@ -319,11 +319,13 @@ fn genTable() Table {
 
         // events
         single(&result, 0x19, source, source, .ignore);
-        range(&result, 0, 0x17, source, source, .ignore);
+        range(&result, 0, 0x06, source, source, .ignore);
+        range(&result, 0x08, 0x17, source, source, .ignore);
         range(&result, 0x1C, 0x1F, source, source, .ignore);
         range(&result, 0x20, 0x7F, source, source, .osc_put);
 
         // => ground
+        single(&result, 0x07, source, .ground, .none);
         single(&result, 0x9C, source, .ground, .none);
     }