path: root/mupdf.zig

const std = @import("std");

/// MuPDF 1.27.0's release archive does not publish a Zig build. Keep its
/// upstream source-of-truth (`Makelists`) and compile the selected C sources
/// as an ordinary Zig static library. This is intentionally a PDF-only first
/// slice: the application needs rendering and structured text, not MuPDF's
/// other document handlers, tools, Javascript engine, OCR, or writers.
pub const Result = struct {
    dependency: *std.Build.Dependency,
    library: *std.Build.Step.Compile,
    /// Whether this archive carries the JPEG 2000 decoder. Remembered rather
    /// than re-derived because `linkTo` has to hand consumers the SAME
    /// preprocessor view the archive was built with — `FZ_ENABLE_JPX` reaches
    /// the public headers, and a consumer that disagrees is an ODR bug that
    /// only shows up as a wrong struct layout at runtime.
    jpx: bool,

    /// Give a Zig module the same preprocessor view as the library, expose the
    /// public C headers to @cImport, and propagate the static link.
    pub fn linkTo(self: Result, module: *std.Build.Module) void {
        configureModule(module, self.dependency, self.jpx);
        // MuPDF's public context headers select the lock-debugging contract
        // from NDEBUG. Every consumer must therefore agree with the archive,
        // even when its own Zig optimization/safety mode differs.
        module.addCMacro("NDEBUG", "1");
        module.linkLibrary(self.library);
    }
};

pub fn add(b: *std.Build, io: std.Io, dependency: *std.Build.Dependency, options: struct {
    target: std.Build.ResolvedTarget,
    optimize: std.builtin.OptimizeMode,
    /// Compile openjpeg and let MuPDF decode JPEG 2000. See the OPENJPEG
    /// block below for why this is a switch rather than always-on.
    jpx: bool = true,
}) Result {
    const module = b.createModule(.{
        .target = options.target,
        .optimize = options.optimize,
        .link_libc = true,
    });
    configureModule(module, dependency, options.jpx);
    // MuPDF's upstream release build defines NDEBUG independently of the
    // optimizer. Without it, an optimized Zig build still enables Fitz's
    // debug-locking/assertion paths. linkTo applies the same public-header
    // contract to consumers; keeping it here also covers every archive source.
    module.addCMacro("NDEBUG", "1");

    // MuPDF itself uses wildcards for these two directories. The dependency
    // is immutable and content-addressed, so discovering the flat C lists at
    // configure time is deterministic; sorting prevents filesystem order from
    // perturbing Zig's cache key.
    module.addCSourceFiles(.{
        .root = dependency.path(""),
        .files = flatCSources(b, io, dependency, "source/fitz"),
        // MuPDF's C uses intentional patterns that Zig/Clang's trap-mode
        // UBSan rejects in ReleaseSafe (opening any document traps in
        // fz_drop_archive). Keep ReleaseSafe's optimization while leaving
        // undefined-behaviour instrumentation off for this upstream layer.
        .flags = &.{ "-std=gnu11", "-fno-sanitize=undefined" },
    });
    module.addCSourceFiles(.{
        .root = dependency.path(""),
        .files = flatCSources(b, io, dependency, "source/pdf"),
        .flags = &.{ "-std=gnu11", "-fno-sanitize=undefined" },
    });

    // TOFU + TOFU_CJK leaves the Base-14 PDF fonts enabled. The official
    // source release ships their generated C blobs, so no host-side objcopy or
    // shell generator is involved.
    module.addCSourceFiles(.{
        .root = dependency.path(""),
        .files = flatCSources(b, io, dependency, "generated/resources/fonts/urw"),
        // These are large byte arrays; optimization only wastes build time.
        // This matches MuPDF's own generated-resource compilation rule.
        .flags = &.{ "-std=gnu11", "-O0" },
    });

    // Parse only the upstream variables we actually select. These lists are
    // the same ones MuPDF's Makefile places in libmupdf-third.a; keeping the
    // names in Makelists avoids a large, easy-to-stale duplicate in this file.
    const makelists = readMakelists(b, io, dependency);
    module.addCSourceFiles(.{
        .root = dependency.path(""),
        .files = makeSources(b, makelists, "FREETYPE_SRC"),
        .flags = &.{
            "-std=gnu11",
            "-DFT_CONFIG_MODULES_H=\"slimftmodules.h\"",
            "-DFT_CONFIG_OPTIONS_H=\"slimftoptions.h\"",
            "-DFT2_BUILD_LIBRARY",
        },
    });
    module.addCSourceFiles(.{
        .root = dependency.path(""),
        .files = makeSources(b, makelists, "LIBJPEG_SRC"),
        .flags = &.{"-std=gnu11"},
    });
    module.addCSourceFiles(.{
        .root = dependency.path(""),
        .files = makeSources(b, makelists, "ZLIB_SRC"),
        .flags = &.{ "-std=gnu11", "-DHAVE_UNISTD_H", "-DHAVE_STDARG_H" },
    });
    module.addCSourceFiles(.{
        .root = dependency.path(""),
        .files = makeSources(b, makelists, "JBIG2DEC_SRC"),
        .flags = &.{
            "-std=gnu11",
            "-DHAVE_STDINT_H",
            "-DJBIG_EXTERNAL_MEMENTO_H=\"mupdf/memento.h\"",
        },
    });

    // --- OPENJPEG ---
    //
    // The JPEG 2000 decoder, and the reason a scanned PDF is not a blank page:
    // a scan is typically one /JPXDecode image per page, so without this MuPDF
    // raises "JPX support disabled" for every one of them and hands back a
    // page with nothing drawn on it. Everything else still works — the pane
    // opens, the page count is right — which makes it look like a renderer bug
    // rather than a missing codec.
    //
    // A switch rather than always-on because it is 31 files of third-party C
    // that exists to parse untrusted input, and openjpeg has the CVE history
    // to match. Default on: a PDF viewer that cannot open scans is the more
    // surprising default, and the ones that matter are exactly the documents
    // nobody can retypeset.
    //
    // Flags are MuPDF's own OPENJPEG_CFLAGS plus OPENJPEG_BUILD_CFLAGS from
    // Makelists; the include path is already on the module (configureModule
    // adds it unconditionally, because encode-jpx.c wants the header even when
    // the codec is off). -fno-sanitize=undefined for the reason source/fitz
    // needs it: this is upstream C full of deliberate wrapping arithmetic, and
    // ReleaseSafe's trap-mode UBSan would turn a legal shift into a crash.
    if (options.jpx) module.addCSourceFiles(.{
        .root = dependency.path(""),
        .files = makeSources(b, makelists, "OPENJPEG_SRC"),
        .flags = &.{
            "-std=gnu11",
            "-fno-sanitize=undefined",
            "-DOPJ_STATIC",
            "-DOPJ_HAVE_INTTYPES_H",
            "-DOPJ_HAVE_STDINT_H",
            "-DMUTEX_pthread=0",
        },
    });

    const library = b.addLibrary(.{
        .name = "mupdf",
        .linkage = .static,
        .root_module = module,
    });
    if (options.target.result.os.tag != .windows)
        module.linkSystemLibrary("m", .{});

    return .{ .dependency = dependency, .library = library, .jpx = options.jpx };
}

/// Add a link-complete smoke test. Unlike merely producing an archive, this
/// exercises document registration, page loading, rasterization and search,
/// so an omitted source or codec symbol fails the build before application
/// integration starts.
pub fn addProbe(b: *std.Build, result: Result, options: struct {
    target: std.Build.ResolvedTarget,
    pdf: std.Build.LazyPath,
}) *std.Build.Step {
    const generated = b.addWriteFiles();
    const source = generated.add("mupdf-probe.c",
        \\#include <mupdf/fitz.h>
        \\#include <stdio.h>
        \\
        \\int main(int argc, char **argv)
        \\{
        \\    fz_context *ctx = NULL;
        \\    fz_document *doc = NULL;
        \\    fz_page *page = NULL;
        \\    fz_pixmap *pix = NULL;
        \\    fz_stext_page *text = NULL;
        \\    fz_search *search = NULL;
        \\    fz_search_result search_result;
        \\    int pages = 0, hit_count = 0, failed = 0;
        \\
        \\    if (argc != 2) return 2;
        \\    ctx = fz_new_context(NULL, NULL, FZ_STORE_DEFAULT);
        \\    if (!ctx) return 3;
        \\    fz_var(doc);
        \\    fz_var(page);
        \\    fz_var(pix);
        \\    fz_var(text);
        \\    fz_var(search);
        \\    fz_try(ctx)
        \\    {
        \\        fz_register_document_handlers(ctx);
        \\        doc = fz_open_document(ctx, argv[1]);
        \\        pages = fz_count_pages(ctx, doc);
        \\        if (pages < 1) fz_throw(ctx, FZ_ERROR_FORMAT, "empty document");
        \\        page = fz_load_page(ctx, doc, 0);
        \\        pix = fz_new_pixmap_from_page(ctx, page, fz_scale(0.25f, 0.25f), fz_device_rgb(ctx), 0);
        \\        text = fz_new_stext_page_from_page(ctx, page, NULL);
        \\        search = fz_new_search(ctx);
        \\        fz_search_set_options(ctx, search, FZ_SEARCH_IGNORE_CASE, "Pardes");
        \\        fz_feed_search(ctx, search, fz_keep_stext_page(ctx, text), 0);
        \\        for (;;)
        \\        {
        \\            search_result = fz_search_forwards(ctx, search);
        \\            if (search_result.reason == FZ_SEARCH_COMPLETE) break;
        \\            if (search_result.reason == FZ_SEARCH_MORE_INPUT)
        \\                fz_feed_search(ctx, search, NULL, search_result.u.more_input.seq_needed);
        \\            else if (search_result.reason == FZ_SEARCH_MATCH)
        \\                ++hit_count;
        \\            else
        \\                fz_throw(ctx, FZ_ERROR_FORMAT, "invalid iterative search state");
        \\        }
        \\        if (!pix || pix->w < 1 || pix->h < 1 || hit_count < 1)
        \\            fz_throw(ctx, FZ_ERROR_FORMAT, "render/search probe failed");
        \\        printf("MuPDF probe: %d pages, %d hit(s), first page %dx%d pixels\n", pages, hit_count, pix->w, pix->h);
        \\    }
        \\    fz_catch(ctx)
        \\    {
        \\        fz_report_error(ctx);
        \\        failed = 1;
        \\    }
        \\    fz_drop_search(ctx, search);
        \\    fz_drop_stext_page(ctx, text);
        \\    fz_drop_pixmap(ctx, pix);
        \\    fz_drop_page(ctx, page);
        \\    fz_drop_document(ctx, doc);
        \\    fz_drop_context(ctx);
        \\    return failed;
        \\}
    );

    const probe = b.addExecutable(.{
        .name = "pardes-mupdf-probe",
        .root_module = b.createModule(.{
            .target = options.target,
            .optimize = .ReleaseSafe,
            .link_libc = true,
        }),
    });
    result.linkTo(probe.root_module);
    probe.root_module.addCSourceFile(.{ .file = source, .flags = &.{"-std=gnu11"} });

    const run = b.addRunArtifact(probe);
    run.addFileArg(options.pdf);
    run.expectExitCode(0);
    return &run.step;
}

fn configureModule(module: *std.Build.Module, dependency: *std.Build.Dependency, jpx: bool) void {
    module.addIncludePath(dependency.path("include"));
    module.addIncludePath(dependency.path("source/fitz"));
    module.addIncludePath(dependency.path("source/pdf"));
    module.addIncludePath(dependency.path("thirdparty/freetype/include"));
    module.addIncludePath(dependency.path("scripts/freetype"));
    module.addIncludePath(dependency.path("thirdparty/libjpeg"));
    module.addIncludePath(dependency.path("scripts/libjpeg"));
    module.addIncludePath(dependency.path("thirdparty/zlib"));
    module.addIncludePath(dependency.path("thirdparty/jbig2dec"));
    // encode-jpx.c includes this header before its FZ_ENABLE_JPX guard.
    module.addIncludePath(dependency.path("thirdparty/openjpeg/src/lib/openjp2"));

    inline for (.{
        .{ "FZ_ENABLE_PDF", "1" },
        .{ "FZ_ENABLE_XPS", "0" },
        .{ "FZ_ENABLE_SVG", "0" },
        .{ "FZ_ENABLE_CBZ", "0" },
        .{ "FZ_ENABLE_IMG", "0" },
        .{ "FZ_ENABLE_HTML", "0" },
        .{ "FZ_ENABLE_FB2", "0" },
        .{ "FZ_ENABLE_MOBI", "0" },
        .{ "FZ_ENABLE_EPUB", "0" },
        .{ "FZ_ENABLE_OFFICE", "0" },
        .{ "FZ_ENABLE_TXT", "0" },
        .{ "FZ_ENABLE_HTML_ENGINE", "0" },
        .{ "FZ_ENABLE_OCR_OUTPUT", "0" },
        .{ "FZ_ENABLE_ODT_OUTPUT", "0" },
        .{ "FZ_ENABLE_DOCX_OUTPUT", "0" },
        .{ "FZ_ENABLE_JS", "0" },
        .{ "FZ_ENABLE_BARCODE", "0" },
        .{ "FZ_ENABLE_BROTLI", "0" },
        .{ "FZ_ENABLE_ICC", "0" },
    }) |macro| module.addCMacro(macro[0], macro[1]);
    // The one codec that is a build option rather than a fixed answer, and it
    // reaches the public headers — so consumers get it through linkTo from the
    // archive's own `jpx`, never re-derived. See the OPENJPEG block in add().
    module.addCMacro("FZ_ENABLE_JPX", if (jpx) "1" else "0");
    module.addCMacro("OCR_DISABLED", "1");
    module.addCMacro("TOFU", "1");
    module.addCMacro("TOFU_CJK", "1");
}

fn flatCSources(b: *std.Build, io: std.Io, dependency: *std.Build.Dependency, relative_dir: []const u8) []const []const u8 {
    var dir = std.Io.Dir.cwd().openDir(io, dependency.path(relative_dir).getPath(b), .{ .iterate = true }) catch
        std.debug.panic("MuPDF source directory is missing: {s}", .{relative_dir});
    defer dir.close(io);

    const max_flat_sources = 4096;
    var bounded: [max_flat_sources][]const u8 = undefined;
    var count: usize = 0;
    var it = dir.iterate();
    while (it.next(io) catch @panic("read MuPDF source directory")) |entry| {
        // Some filesystems report d_type as unknown. The selected directories
        // are flat and immutable, so the suffix is the reliable predicate.
        if (!std.mem.endsWith(u8, entry.name, ".c")) continue;
        if (count == bounded.len) std.debug.panic("MuPDF source directory exceeds {d} C files: {s}", .{ bounded.len, relative_dir });
        bounded[count] = b.fmt("{s}/{s}", .{ relative_dir, entry.name });
        count += 1;
    }
    if (count == 0) std.debug.panic("MuPDF source directory has no C files: {s}", .{relative_dir});
    const files = b.allocator.alloc([]const u8, count) catch @panic("OOM");
    @memcpy(files, bounded[0..count]);
    std.mem.sort([]const u8, files, {}, lessThan);
    return files;
}

fn readMakelists(b: *std.Build, io: std.Io, dependency: *std.Build.Dependency) []const u8 {
    return std.Io.Dir.cwd().readFileAlloc(
        io,
        dependency.path("Makelists").getPath(b),
        b.allocator,
        .limited(1024 * 1024),
    ) catch @panic("read MuPDF Makelists");
}

fn makeSources(b: *std.Build, makelists: []const u8, variable: []const u8) []const []const u8 {
    var count: usize = 0;
    var lines = std.mem.splitScalar(u8, makelists, '\n');
    while (lines.next()) |raw_line| {
        const line = std.mem.trim(u8, raw_line, " \t\r");
        if (!std.mem.startsWith(u8, line, variable)) continue;
        const assignment = std.mem.trimStart(u8, line[variable.len..], " \t");
        if (!std.mem.startsWith(u8, assignment, "+=")) continue;
        const path = std.mem.trim(u8, assignment[2..], " \t\r");
        if (std.mem.endsWith(u8, path, ".c")) count += 1;
    }
    if (count == 0) std.debug.panic("MuPDF Makelists variable has no C sources: {s}", .{variable});
    const files = b.allocator.alloc([]const u8, count) catch @panic("OOM");
    lines = std.mem.splitScalar(u8, makelists, '\n');
    var file_i: usize = 0;
    while (lines.next()) |raw_line| {
        const line = std.mem.trim(u8, raw_line, " \t\r");
        if (!std.mem.startsWith(u8, line, variable)) continue;
        const assignment = std.mem.trimStart(u8, line[variable.len..], " \t");
        if (!std.mem.startsWith(u8, assignment, "+=")) continue;
        const path = std.mem.trim(u8, assignment[2..], " \t\r");
        if (!std.mem.endsWith(u8, path, ".c")) continue;
        files[file_i] = b.dupe(path);
        file_i += 1;
    }
    return files;
}

fn lessThan(_: void, a: []const u8, b: []const u8) bool {
    return std.mem.order(u8, a, b) == .lt;
}