commit c4b460894d40e0f78342c2b502b8408b1c2f77bc
Author: Nulo <git@nulo.in>
Date:   Sat Jul 31 23:25:00 2021 +0000

    Init

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..e73c965
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,2 @@
+zig-cache/
+zig-out/
diff --git a/markdown.zig b/markdown.zig
new file mode 100644
index 0000000..382e6e3
--- /dev/null
+++ b/markdown.zig
@@ -0,0 +1,203 @@
+pub const std = @import("std");
+
+pub const Result = struct { result: Block, rest: []const u8 };
+const definedBlockTypes = [_]type{ Heading, CodeBlock };
+const blockTypes = definedBlockTypes ++ [_]type{Paragraph};
+
+pub const Block = union(enum) {
+    paragraph: Paragraph,
+    heading: Heading,
+    code_block: CodeBlock,
+};
+
+pub const Paragraph = struct {
+    text: []const u8,
+
+    pub fn parse(allocator: ?*std.mem.Allocator, input: []const u8) !?Result {
+        _ = allocator;
+        var rest = input;
+        if (rest.len == 0) return null;
+        while (rest.len > 0) : (rest = rest[1..]) {
+            inline for (definedBlockTypes) |BlockType| {
+                if (try BlockType.parse(null, rest)) |_| {
+                    const text = trim(input[0 .. input.len - rest.len]);
+                    if (text.len == 0) return null;
+                    return Result{ .result = .{
+                        .paragraph = .{ .text = text },
+                    }, .rest = rest };
+                }
+            }
+            if (std.mem.startsWith(u8, rest, "\n\n")) {
+                const text = trim(input[0 .. input.len - rest.len]);
+                if (text.len == 0) return null;
+                return Result{ .result = .{
+                    .paragraph = .{ .text = text },
+                }, .rest = rest };
+            }
+        }
+        const text = trim(input);
+        if (text.len == 0) return null;
+        return Result{ .result = .{
+            .paragraph = .{ .text = text },
+        }, .rest = input[input.len..] };
+    }
+};
+
+pub const Heading = struct {
+    text: []const u8,
+    level: Level,
+
+    pub const Level = enum(u3) {
+        one = 1,
+        two = 2,
+        three = 3,
+        four = 4,
+        five = 5,
+        six = 6,
+    };
+
+    pub fn parse(_: ?*std.mem.Allocator, input: []const u8) !?Result {
+        inline for (std.meta.fields(Level)) |level| {
+            const str = ("#" ** level.value) ++ " ";
+            if (std.mem.startsWith(u8, input, str)) {
+                const text = untilNewline(input[str.len..]);
+                return Result{ .result = .{
+                    .heading = .{
+                        .text = text,
+                        .level = @intToEnum(Level, level.value),
+                    },
+                }, .rest = input[str.len + text.len ..] };
+            }
+        }
+        return null;
+    }
+};
+
+pub const CodeBlock = struct {
+    language: []const u8,
+    text: []const u8,
+
+    pub fn parse(_: ?*std.mem.Allocator, input: []const u8) !?Result {
+        if (std.mem.startsWith(u8, input, "```")) {
+            const language = std.mem.sliceTo(input[3..], '\n');
+            const rest = input[3 + language.len + 1 ..];
+            if (std.mem.indexOf(u8, rest, "```")) |end| {
+                return Result{
+                    .result = .{
+                        .code_block = .{
+                            .language = language,
+                            .text = trim(rest[0..end]),
+                        },
+                    },
+                    .rest = rest[end + 3 ..],
+                };
+            }
+        }
+        return null;
+    }
+};
+
+pub fn untilNewline(input: []const u8) []const u8 {
+    return if (std.mem.indexOf(u8, input, "\n")) |index| input[0..index] else input;
+}
+
+/// Returns the index of where the next block starts.
+pub fn untilNextBlock(input: []const u8) usize {
+    for (input) |_, i| {
+        const rest = input[i..];
+        if (std.mem.startsWith(u8, rest, "\n\n") or
+            Heading.parse(rest) != null or
+            CodeBlock.parse(rest) != null)
+        {
+            return i;
+        }
+    }
+    return input.len;
+}
+pub fn trim(input: []const u8) []const u8 {
+    return std.mem.trim(u8, input, " \n");
+}
+
+pub const Tree = std.ArrayList(Block);
+pub fn parse(
+    allocator: *std.mem.Allocator,
+    input: []const u8,
+) !Tree {
+    var tree = Tree.init(allocator);
+    errdefer tree.deinit();
+
+    const orig = trim(input);
+    var rest = orig;
+    while (rest.len > 0) {
+        var ran = false;
+        inline for (blockTypes) |blockType| {
+            if (try blockType.parse(allocator, rest)) |result| {
+                try tree.append(result.result);
+                rest = result.rest;
+                // It would be better to just continue the outer while loop but
+                // that makes the compiler crash.
+                ran = true;
+                break;
+            }
+        }
+        if (!ran) rest = rest[1..];
+    }
+    return tree;
+}
+
+pub fn main() !void {
+    var gpa = std.heap.GeneralPurposeAllocator(.{}){};
+    defer _ = gpa.deinit();
+    const allocator = &gpa.allocator;
+
+    const input = try std.io.getStdIn().readToEndAlloc(allocator, 10241024);
+    defer allocator.free(input);
+    const stdout = std.io.getStdOut().writer();
+
+    const parsed = try parse(allocator, input);
+    defer parsed.deinit();
+
+    for (parsed.items) |block| {
+        switch (block) {
+            .paragraph => |p| try stdout.print("<p>{s}</p>", .{p.text}),
+            .heading => |h| try stdout.print("<h{0}>{1s}</h{0}>", .{ @enumToInt(h.level), h.text }),
+            .code_block => |c| try stdout.print("<pre><code>{s}</code></pre>", .{c.text}),
+        }
+    }
+}
+
+test "parse headings and paragraph" {
+    const allocator = std.testing.allocator;
+    const input =
+        \\# Hola
+        \\## Hola
+        \\
+        \\### Heey
+        \\
+        \\:)
+        \\## Hola mundo
+    ;
+
+    const parsed = try parse(allocator, input);
+    defer parsed.deinit();
+
+    try std.testing.expectEqual(@as(usize, 5), parsed.items.len);
+
+    try std.testing.expectEqualStrings("heading", @tagName(parsed.items[0]));
+    try std.testing.expect(parsed.items[0].heading.level == .one);
+    try std.testing.expectEqualStrings("Hola", parsed.items[0].heading.text);
+
+    try std.testing.expectEqualStrings("heading", @tagName(parsed.items[1]));
+    try std.testing.expect(parsed.items[1].heading.level == .two);
+    try std.testing.expectEqualStrings("Hola", parsed.items[1].heading.text);
+
+    try std.testing.expectEqualStrings("heading", @tagName(parsed.items[2]));
+    try std.testing.expect(parsed.items[2].heading.level == .three);
+    try std.testing.expectEqualStrings("Heey", parsed.items[2].heading.text);
+
+    try std.testing.expectEqualStrings("paragraph", @tagName(parsed.items[3]));
+    try std.testing.expectEqualStrings(":)", parsed.items[3].paragraph.text);
+
+    try std.testing.expectEqualStrings("heading", @tagName(parsed.items[4]));
+    try std.testing.expectEqualStrings("Hola mundo", parsed.items[4].heading.text);
+}