Conservancy/gpl-compliance-tools
6
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Begin work on processing maildir files.

Browse source
This commit is contained in:
Bradley M. Kuhn 2023-06-02 21:48:27 -07:00
parent c845c825dd
commit 123dbd6018
1 changed files with 44 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

45
build-label-upi-number.plx
View file

@ -16,6 +16,9 @@ use Getopt::Long;
use File::Spec::Functions;
use File::Spec;
use File::Path qw(make_path);
use Mail::Header;
use Email::Address::XS;
use File::Copy;
my %GROUP_NAMES_BY_DIR = ( confidential => 'CONFIDENTIAL', privilege => 'PRIVILEGE', privileged => 'PRIVILEGE',
'journalist-privilege' => 'PRIVILEGE' );
@ -61,6 +64,36 @@ UsageAndExit("Error reading \"$upiNumberFile\"") unless $count == 1 and $upiStar
my $upiCurrentNum = $upiStart;
sub ProcessMailDir($$) {
my($inputMailDir, $destOutDir) = @_;
my @msgDirs = (catfile($inputMailDir, 'new'), catfile($inputMailDir, 'cur'));
foreach my $dir (@msgDirs) {
die "$inputMailDir must be a readble maildir folder but $dir isn't a readable directory: $!"
unless (-d $inputMailDir);
}
foreach my $dir (@msgDirs) {
opendir(my $mailDirFH, $dir);
while (my $file = readdir $mailDirFH) {
next if -d $file; # skip directories
my $msgFile = catfile($dir, $file);
open(my $msgFH, "<", $msgFile);
print " $msgFile\n";
my $header = new Mail::Header($msgFH);
my $fields = $header->header_hashref;
foreach my $fieldName (qw/From To Cc Subject Date/) {
foreach my $item (@{$fields->{$fieldName}}) {
chomp $item;
print " $fieldName $item\n";
}
}
close $msgFH;
}
closedir $mailDirFH;
}
}
# Main Loop:
opendir(my $topDH, $INPUT_TOPLEVEL_DIR);
while (my $rfp = readdir $topDH) {
next unless $rfp =~ /^\s*rfp-(\d+)\s*$/;
@ -69,7 +102,17 @@ while (my $rfp = readdir $topDH) {
while (my $bucketName = readdir $rfpDH) {
next unless $bucketName eq $GROUP;
my $inBucketDirName = catfile($INPUT_TOPLEVEL_DIR, $rfp, $bucketName);
print " $inBucketDirName\n";
opendir(my $bucketDH, $inBucketDirName);
while (my $typeName = readdir $bucketDH) {
next if $typeName =~ /^\s*\.\.?\s*$/;
my $typeDirName = catfile($INPUT_TOPLEVEL_DIR, $rfp, $bucketName, $typeName);
die "regular file found where we expected a type in $typeName" unless -d $typeDirName;
print " $typeDirName\n";
if ($typeName =~ /email/i) {
ProcessMailDir($typeDirName, "");
}
}
closedir $bucketDH;
}
closedir $rfpDH;
}