1/jecolon/zigstr v0.20

A UTF-8 string type for Zig.


A UTF-8 string type.

What? No Characters?

Zigstr tries to emphasize the clear distinction between bytes (u8), code points (u21), and grapheme clusters ([]const u8) as per the Unicode standard. Note that the term character is glaringly missing here, as it tends to produce more confusion than clarity, and in fact Unicode has no concrete character concept, only abstract characters are broadly mentioned. The closest concept resembling a human-perceivable character in Unicode is the Grapheme Cluster, represented here as the Grapheme type returned from each call to the next method on a GraphemeIterator (see sample code below).


There are two possibilities when creating a new Zigstr:

  • You own the bytes, requiring the deinit method to free them.
  • You don't own the bytes, so deinit will not free them.

To create a Zigstr in each of these circumstances:

// Here the slice is []const u8.
var str = try Zigstr.fromBytes(allocator, "Hello");
defer str.deinit(); // still need `deinit` to free other resources, but not the passed-in bytes.

// Here the slice is []u8.
var str = try Zigstr.fromOwnedBytes(allocator, slice);
defer str.deinit(); // owned bytes will be freed.

String Comparison, Sorting (Collation), Normalization, Display With, and More

Given the large amounts of data required for many Unicode string operations, these operations are not included in the Zigstr struct to keep it light and fast. For the myriad Unicode text processing operations, check out the Ziglyph library.

Integrating Zigstr in your Project

Zigstr uses Zigmod for dependency management. It's the easiest way to include this library in your project. Once you have Zigmod, you just have to run:

$ zigmod aq add 1/jecolon/zigstr
$ zigmod fetch

Now in your build.zig you add this import:

const deps = @import("deps.zig");

In the exe section for the executable where you wish to have Zigstr available, add:


Now in the code, you can import Zigstr like this:

const zigstr = @import("zigstr");

Usage Examples

const zigstr = @import("zigstr");

test "Zigstr README tests" {
    var allocator = std.testing.allocator;
    var str = try zigstr.fromBytes(std.testing.allocator, "Héllo");
    defer str.deinit();

    // Byte count.
    try expectEqual(@as(usize, 6), str.byteCount());

    // Code point iteration.
    var cp_iter = str.codePointIter();
    var want = [_]u21{ 'H', 0x00E9, 'l', 'l', 'o' };

    var i: usize = 0;
    while (cp_iter.next()) |cp| : (i += 1) {
        try expectEqual(want[i], cp.scalar);

    // Code point count.
    try expectEqual(@as(usize, 5), try str.codePointCount());

    // Collect all code points at once.
    const code_points = try str.codePoints(allocator);
    defer allocator.free(code_points);
    for (code_points) |cp, j| {
        try expectEqual(want[j], cp);

    // Grapheme cluster iteration.
    var giter = try str.graphemeIter();
    const gc_want = [_][]const u8{ "H", "é", "l", "l", "o" };

    i = 0;
    while (giter.next()) |gc| : (i += 1) {
        try expect(gc.eql(gc_want[i]));

    // Collect all grapheme clusters at once.
    try expectEqual(@as(usize, 5), try str.graphemeCount());
    const gcs = try str.graphemes(allocator);
    defer allocator.free(gcs);

    for (gcs) |gc, j| {
        try expect(gc.eql(gc_want[j]));

    // Grapheme count.
    try expectEqual(@as(usize, 5), try str.graphemeCount());

    // Indexing (with negative indexes too.)
    try expectEqual(try str.byteAt(0), 72); // H
    try expectEqual(try str.byteAt(-2), 108); // l
    try expectEqual(try str.codePointAt(0), 'H');
    try expectEqual(try str.codePointAt(-2), 'l');
    try expect((try str.graphemeAt(0)).eql("H"));
    try expect((try str.graphemeAt(-4)).eql("é"));

    // Copy
    var str2 = try str.copy(allocator);
    defer str2.deinit();
    try expect(str.eql(str2.bytes.items));
    try expect(str2.eql("Héllo"));
    try expect(str.sameAs(str2));

    // Empty and obtain owned slice of bytes.
    const bytes2 = try str2.toOwnedSlice();
    defer allocator.free(bytes2);
    try expect(str2.eql(""));
    try expectEqualStrings(bytes2, "Héllo");

    // Re-initialize a Zigstr.
    try str.reset("foo");

    // Equality
    try expect(str.eql("foo")); // exact
    try expect(!str.eql("fooo")); // lengths
    try expect(!str.eql("foó")); // combining marks
    try expect(!str.eql("Foo")); // letter case

    // Trimming.
    try str.reset("   Hello");
    try str.trimLeft(" ");
    try expect(str.eql("Hello"));

    try str.reset("Hello   ");
    try str.trimRight(" ");
    try expect(str.eql("Hello"));

    try str.reset("   Hello   ");
    try str.trim(" ");
    try expect(str.eql("Hello"));

    // indexOf / contains / lastIndexOf
    try str.reset("H\u{65}\u{301}llo"); // Héllo
    try expectEqual(try str.indexOf("l"), 2);
    try expectEqual(try str.indexOf("z"), null);
    try expect(try str.contains("l"));
    try expect(!try str.contains("z"));
    try expectEqual(try str.lastIndexOf("l"), 3);
    try expectEqual(try str.lastIndexOf("z"), null);

    // count
    try expectEqual(str.count("l"), 2);
    try expectEqual(str.count("ll"), 1);
    try expectEqual(str.count("z"), 0);

    // Tokenization
    try str.reset(" Hello World ");

    // Token iteration.
    var tok_iter = str.tokenIter(" ");
    try expectEqualStrings("Hello", tok_iter.next().?);
    try expectEqualStrings("World", tok_iter.next().?);
    try expect(tok_iter.next() == null);

    // Collect all tokens at once.
    var ts = try str.tokenize(" ", allocator);
    defer allocator.free(ts);
    try expectEqual(@as(usize, 2), ts.len);
    try expectEqualStrings("Hello", ts[0]);
    try expectEqualStrings("World", ts[1]);

    // Split
    var split_iter = str.splitIter(" ");
    try expectEqualStrings("", split_iter.next().?);
    try expectEqualStrings("Hello", split_iter.next().?);
    try expectEqualStrings("World", split_iter.next().?);
    try expectEqualStrings("", split_iter.next().?);
    try expect(split_iter.next() == null);

    // Collect all sub-strings at once.
    var ss = try str.split(" ", allocator);
    defer allocator.free(ss);
    try expectEqual(@as(usize, 4), ss.len);
    try expectEqualStrings("", ss[0]);
    try expectEqualStrings("Hello", ss[1]);
    try expectEqualStrings("World", ss[2]);
    try expectEqualStrings("", ss[3]);

    // Convenience methods for splitting on newline '\n'.
    try str.reset("Hello\nWorld");
    var iter = str.lineIter(); // line iterator
    try expectEqualStrings(iter.next().?, "Hello");
    try expectEqualStrings(iter.next().?, "World");

    var lines_array = try str.lines(allocator); // array of lines without ending \n.
    defer allocator.free(lines_array);
    try expectEqualStrings(lines_array[0], "Hello");
    try expectEqualStrings(lines_array[1], "World");

    // startsWith / endsWith
    try str.reset("Hello World");
    try expect(str.startsWith("Hell"));
    try expect(!str.startsWith("Zig"));
    try expect(str.endsWith("World"));
    try expect(!str.endsWith("Zig"));

    // Concatenation
    try str.reset("Hello");
    try str.concat(" World");
    try expect(str.eql("Hello World"));
    var others = [_][]const u8{ " is", " the", " tradition!" };
    try str.concatAll(&others);
    try expect(str.eql("Hello World is the tradition!"));

    // replace
    try str.reset("Hello");
    var replacements = try str.replace("l", "z");
    try expectEqual(@as(usize, 2), replacements);
    try expect(str.eql("Hezzo"));

    replacements = try str.replace("z", "");
    try expectEqual(@as(usize, 2), replacements);
    try expect(str.eql("Heo"));

    // Append a code point or many.
    try str.reset("Hell");
    try str.append('o');
    try expectEqual(@as(usize, 5), str.bytes.items.len);
    try expect(str.eql("Hello"));
    try str.appendAll(&[_]u21{ ' ', 'W', 'o', 'r', 'l', 'd' });
    try expect(str.eql("Hello World"));

    // Test for empty string.
    try expect(!str.isEmpty());

    // Test for whitespace only (blank) strings.
    try str.reset("  \t  ");
    try expect(try str.isBlank());
    try expect(!str.isEmpty());

    // Remove grapheme clusters (characters) from strings.
    try str.reset("Hello World");
    try str.dropLeft(6);
    try expect(str.eql("World"));
    try str.reset("Hello World");
    try str.dropRight(6);
    try expect(str.eql("Hello"));

    // Insert at a grapheme index.
    try str.insert("Hi", 0);
    try expect(str.eql("HiHello"));

    // Remove a sub-string.
    try str.remove("Hi");
    try expect(str.eql("Hello"));
    try str.remove("Hello");
    try expect(str.eql(""));

    // Repeat a string's content.
    try str.reset("*");
    try str.repeat(10);
    try expect(str.eql("**********"));
    try str.repeat(1);
    try expect(str.eql("**********"));
    try str.repeat(0);
    try expect(str.eql(""));

    // Reverse a string. Note correct handling of Unicode code point ordering.
    try str.reset("Héllo 😊");
    try str.reverse();
    try expect(str.eql("😊 olléH"));

    // You can also construct a Zigstr from coce points.
    const cp_array = [_]u21{ 0x68, 0x65, 0x6C, 0x6C, 0x6F }; // "hello"
    str = try zigstr.fromCodePoints(allocator, &cp_array);
    try expect(str.eql("hello"));
    try expectEqual(str.codePointCount(), 5);

    // Also create a Zigstr from a slice of strings.
    str = try zigstr.fromJoined(std.testing.allocator, &[_][]const u8{ "Hello", "World" }, " ");
    try expect(str.eql("Hello World"));

    // Chomp line breaks.
    try str.reset("Hello\n");
    try str.chomp();
    try expectEqual(@as(usize, 5), str.bytes.items.len);
    try expect(str.eql("Hello"));

    try str.reset("Hello\r");
    try str.chomp();
    try expectEqual(@as(usize, 5), str.bytes.items.len);
    try expect(str.eql("Hello"));

    try str.reset("Hello\r\n");
    try str.chomp();
    try expectEqual(@as(usize, 5), str.bytes.items.len);
    try expect(str.eql("Hello"));

    // byteSlice, codePointSlice, graphemeSlice, substr
    try str.reset("H\u{0065}\u{0301}llo"); // Héllo
    const bytes = try str.byteSlice(1, 4);
    try expectEqualSlices(u8, bytes, "\u{0065}\u{0301}");

    const cps = try str.codePointSlice(allocator, 1, 3);
    defer allocator.free(cps);
    try expectEqualSlices(u21, cps, &[_]u21{ '\u{0065}', '\u{0301}' });

    const gs = try str.graphemeSlice(allocator, 1, 2);
    defer allocator.free(gs);
    try expect(gs[0].eql("\u{0065}\u{0301}"));

    // Substrings
    var sub = try str.substr(1, 2);
    try expectEqualStrings("\u{0065}\u{0301}", sub);

    try expectEqualStrings(bytes, sub);

    // Letter case detection.
    try str.reset("hello! 123");
    try expect(try str.isLower());
    try expect(!try str.isUpper());
    try str.reset("HELLO! 123");
    try expect(try str.isUpper());
    try expect(!try str.isLower());

    // Letter case conversion.
    try str.reset("Héllo World! 123\n");
    try str.toLower();
    try expect(str.eql("héllo world! 123\n"));
    try str.toUpper();
    try expect(str.eql("HÉLLO WORLD! 123\n"));
    try str.reset("tHe (mOviE) 2112\n");
    try str.toTitle();
    try expect(str.eql("The (Movie) 2112\n"));

    // Parsing content.
    try str.reset("123");
    try expectEqual(try str.parseInt(u8, 10), 123);
    try str.reset("123.456");
    try expectEqual(try str.parseFloat(f32), 123.456);
    try str.reset("true");
    try expect(try str.parseBool());

    // Truthy == True, T, Yes, Y, On in any letter case.
    // Not Truthy == False, F, No, N, Off in any letter case.
    try expect(try str.parseTruthy());
    try str.reset("TRUE");
    try expect(try str.parseTruthy());
    try str.reset("T");
    try expect(try str.parseTruthy());
    try str.reset("No");
    try expect(!try str.parseTruthy());
    try str.reset("off");
    try expect(!try str.parseTruthy());

    // Zigstr implements the std.fmt.format interface.
    std.debug.print("Zigstr: {}\n", .{str});

Package Contents

  • .gitattributes
  • zigmod.lock
  • build.zig
  • src/ascii.zig
  • src/main.zig
  • src/Zigstr.zig
  • README.md
  • zig.mod
  • .gitignore


  • git https://github.com/jecolon/ziglyph.git

Root Dependencies

  • git https://github.com/jecolon/ziglyph.git

Build Dependencies

  • git https://github.com/jecolon/ziglyph.git


Published On Tree @ Commit Size
v0.20 Mon, 13 Jun 2022 00:14:21 UTC Tree 73.980 KB
v0.19 Sun, 26 Dec 2021 13:11:19 UTC Tree 74.066 KB
v0.18 Thu, 23 Dec 2021 20:21:34 UTC Tree 74.321 KB
v0.17 Thu, 23 Dec 2021 16:09:49 UTC Tree 73.703 KB
v0.16 Wed, 24 Nov 2021 20:11:11 UTC Tree 73.717 KB
v0.15 Mon, 20 Sep 2021 10:56:03 UTC Tree 73.687 KB
v0.14 Fri, 17 Sep 2021 12:24:16 UTC Tree 73.708 KB
v0.13 Fri, 17 Sep 2021 12:05:21 UTC Tree 73.522 KB
v0.12 Wed, 01 Sep 2021 13:10:29 UTC Tree 73.522 KB
v0.11 Thu, 26 Aug 2021 16:03:46 UTC Tree 73.846 KB
v0.10 Thu, 26 Aug 2021 15:54:28 UTC Tree 73.840 KB
v0.9 Wed, 25 Aug 2021 23:50:30 UTC Tree 73.857 KB
v0.8 Wed, 25 Aug 2021 12:32:11 UTC Tree 73.870 KB
v0.7 Wed, 25 Aug 2021 12:22:59 UTC Tree 74.300 KB
v0.6 Mon, 23 Aug 2021 10:52:32 UTC Tree 73.938 KB
v0.5 Mon, 23 Aug 2021 10:33:51 UTC Tree 73.936 KB
v0.4 Mon, 23 Aug 2021 00:40:59 UTC Tree 73.988 KB
v0.3 Mon, 23 Aug 2021 00:28:15 UTC Tree 73.988 KB
v0.2 Mon, 23 Aug 2021 00:09:15 UTC Tree 74.133 KB
v0.1 Sun, 22 Aug 2021 23:52:08 UTC Tree 74.070 KB