From 9d51df02c110550b81a0e8424089effbb7e9251d Mon Sep 17 00:00:00 2001
From: Yorhel <git@yorhel.nl>
Date: Fri, 15 Nov 2024 11:08:26 +0100
Subject: [PATCH] Add --export-block-size option + minor man page adjustments

---
 ncdu.1             | 34 +++++++++++++++++++++++++++-------
 src/bin_export.zig |  3 ++-
 src/main.zig       |  6 ++++++
 3 files changed, 35 insertions(+), 8 deletions(-)

diff --git a/ncdu.1 b/ncdu.1
index f1e52a7..3f9e341 100644
--- a/ncdu.1
+++ b/ncdu.1
@@ -23,6 +23,7 @@
 .Op Fl t , \-threads Ar num
 .Op Fl c , \-compress , \-no\-compress
 .Op Fl \-compress\-level Ar num
+.Op Fl \-export\-block\-size Ar num
 .Op Fl 0 , 1 , 2
 .Op Fl q , \-slow\-ui\-updates , \-fast\-ui\-updates
 .Op Fl \-enable\-shell , \-disable\-shell
@@ -193,9 +194,18 @@ The binary format (see
 .Fl O )
 does not have this problem and supports efficient exporting with any number of
 threads.
+.El
+.
+.Ss Export Options
+These options affect behavior when exporting to file with the
+.Fl o
+or
+.Fl O
+options.
+.Bl -tag -width Ds
 .It Fl c , \-compress , \-no\-compress
 Enable or disable Zstandard compression when exporting to JSON (see
-.Fl o )
+.Fl o ) .
 .It Fl \-compress\-level Ar num
 Set the Zstandard compression level when using
 .Fl O
@@ -203,6 +213,17 @@ or
 .Fl c .
 Valid values are 1 (fastest) to 19 (slowest).
 Defaults to 4.
+.It Fl \-export\-block\-size Ar num
+Set the block size, in kibibytes, for the binary export format (see
+.Fl O ) .
+Larger blocks require more memory but result in better compression efficiency.
+This option can be combined with a higher
+.Fl \-compress\-level
+for even better compression.
+.Pp
+Accepted values are between 4 and 16000.
+The defaults is to start at 64 KiB and then gradually increase the block size
+for large exports.
 .El
 .
 .Ss Interface Options
@@ -497,29 +518,28 @@ Empty directory.
 .Sh EXAMPLES
 To scan and browse the directory you're currently in, all you need is a simple:
 .Dl ncdu
-If you want to scan a full filesystem, for example your root filesystem, then
-you'll want to use
+To scan a full filesystem, for example your root filesystem, you'll want to use
 .Fl x :
 .Dl ncdu \-x /
 .Pp
 Since scanning a large directory may take a while, you can scan a directory and
 export the results for later viewing:
 .Bd -literal -offset indent
-ncdu \-1cxo export.json.zst /
+ncdu \-1xO export.ncdu /
 # ...some time later:
-ncdu \-f export.json.zst
+ncdu \-f export.ncdu
 .Ed
 To export from a cron job, make sure to replace
 .Fl 1
 with
 .Fl 0
-to suppress any unnecessary output.
+to suppress unnecessary progress output.
 .Pp
 You can also export a directory and browse it once scanning is done:
 .Dl ncdu \-co\- | tee export.json.zst | ./ncdu \-f\-
 .Pp
 To scan a system remotely, but browse through the files locally:
-.Dl ssh user@system ncdu \-co\- / | ./ncdu \-cf\-
+.Dl ssh user@system ncdu \-co\- / | ./ncdu \-f\-
 Remote scanning and local viewing has two major advantages when
 compared to running
 .Nm
diff --git a/src/bin_export.zig b/src/bin_export.zig
index b19628a..43a8cf0 100644
--- a/src/bin_export.zig
+++ b/src/bin_export.zig
@@ -72,7 +72,8 @@ inline fn cborByte(major: CborMajor, arg: u5) u8 { return (@as(u8, @intFromEnum(
 fn blockSize(num: u32) usize {
     //                        block size    uncompressed data in this num range
     //                 # mil      # KiB         # GiB
-    return if (num < ( 1<<20))   64<<10  //    64
+    return main.config.export_block_size
+    orelse if (num < ( 1<<20))   64<<10  //    64
       else if (num < ( 2<<20))  128<<10  //   128
       else if (num < ( 4<<20))  256<<10  //   512
       else if (num < ( 8<<20))  512<<10  //  2048
diff --git a/src/main.zig b/src/main.zig
index 440ec91..2cf0f8e 100644
--- a/src/main.zig
+++ b/src/main.zig
@@ -82,6 +82,7 @@ pub const config = struct {
     pub var threads: usize = 1;
     pub var complevel: u8 = 4;
     pub var compress: bool = false;
+    pub var export_block_size: ?usize = null;
 
     pub var update_delay: u64 = 100*std.time.ns_per_ms;
     pub var scan_ui: ?enum { none, line, full } = null;
@@ -283,6 +284,11 @@ fn argConfig(args: *Args, opt: Args.Option) bool {
         const val = args.arg();
         config.complevel = std.fmt.parseInt(u8, val, 10) catch ui.die("Invalid number for --compress-level: {s}.\n", .{val});
         if (config.complevel <= 0 or config.complevel > 20) ui.die("Invalid number for --compress-level: {s}.\n", .{val});
+    } else if (opt.is("--export-block-size")) {
+        const val = args.arg();
+        const num = std.fmt.parseInt(u14, val, 10) catch ui.die("Invalid number for --export-block-size: {s}.\n", .{val});
+        if (num < 4 or num > 16000) ui.die("Invalid number for --export-block-size: {s}.\n", .{val});
+        config.export_block_size = @as(usize, num) * 1024;
     } else if (opt.is("--confirm-quit")) config.confirm_quit = true
     else if (opt.is("--no-confirm-quit")) config.confirm_quit = false
     else if (opt.is("--confirm-delete")) config.confirm_delete = true