Thank you. I (already) did that but couldn’t find the problem. I tried to create a fancy regex to find unquoted instances of 日本語, but couldn’t figure that out, either.
The tags in many of your journal entries are a mess. I hacked together a quick Perl script to quote each tag and wrap them all in an array, and now the site builds fine for me:
#!/usr/bin/env perl
use File::Find;
use File::Slurp;
find(\&wanted, ".");
sub wanted {
return unless /\.md$/;
my $text = read_file($_);
if ($text =~ /^tags: *(.*)$/m && index($1, "[") == -1) {
write_file($_ . ".old", $text);
my @tags = split(/\s*,\s*/, $1);
my $tags = '[';
$tags .= '"' . join('","', @tags) . '"'
if @tags;
$tags .= ']';
$text =~ s/^tags: *(.*)$/tags: $tags/m;
write_file($_, $text);
}
}
(not every tag needed to be quoted, but even setting aside the Japanese ones, you’ve got some oddballs in there, and given the build time, I wasn’t going to keep iterating to find out exactly which ones needed fixing)
As a side note, the theme you’re using is not designed to handle large amounts of content; it took just under 10 minutes to build under Hugo 0.55.6 with 7,009 MD files in content, where my blog takes about 15 seconds to build successfully with 8,500 MD files:
| EN
+------------------+-------+
Pages | 10095
Paginator pages | 0
Non-page files | 6106
Static files | 35
Processed images | 0
Aliases | 0
Sitemaps | 1
Cleaned | 0
Total in 595760 ms
Double thank you for the note that the theme was not well suited for lots of files. I ended up going back to purehugo, which is fast even with lots of files.
[SOLVED] The original problem in my case was due to not having push/pulled the latest commits of the repo in which I had already cleaned the tags. smh.