Changeset 34921 for main/trunk/greenstone2/perllib/cpan/File/RandomAccess.pod

Timestamp:

2021-02-26T19:39:51+13:00 (3 years ago)

Author:

anupama

Message:

Committing the improvements to EmbeddedMetaPlugin's processing of Keywords vs other metadata fields. Keywords were literally stored as arrays of words rather than phrases in PDFs (at least in Diego's sample PDF), whereas other meta fields like Subjects and Creators stored them as arrays of phrases. To get both to work, Kathy updated EXIF to a newer version, to retrieve the actual EXIF values stored in the PDF. And Kathy and Dr Bainbridge came up with a new option that I added called apply_join_before_split_to_metafields that's a regex which can list the metadata fields to apply the join_before_split to and whcih previously always got applied to all metadata fields. Now it's applied to any *Keywords metafields by default, as that's the metafield we have experience of that behaves differently to the others, as it stores by word instead of phrases. Tested on Diego's sample PDF. Diego has double-checked it to works on his sample PDF too, setting the split char to ; and turning on the join_before_split and leaving apply_join_before_split_to_metafields at its default of .*Keywords. File changes are strings.properties for the tooltip, the plugin introducing the option and working with it and Kathy's EXIF updates affecting cpan/File and cpan/Image.

File:

• main/trunk/greenstone2/perllib/cpan/File/RandomAccess.pod (modified) (3 diffs)

main/trunk/greenstone2/perllib/cpan/File/RandomAccess.pod

@@ -4 +4 @@
 # Description:  Buffer to support random access reading of sequential file
 #
-# Legal:        Copyright (c) 2003-2010 Phil Harvey (phil at owl.phy.queensu.ca)
+# Legal:        Copyright (c) 2003-2021 Phil Harvey (philharvey66 at gmail.com)
 #               This library is free software; you can redistribute it and/or
 #               modify it under the same terms as Perl itself.
@@ -48 +48 @@
 =item Inputs:
 
-0) Reference to RandomAccess object.
+0) Reference to RandomAccess object or RandomAccess class name.
 
 1) File reference or scalar reference.
 
-2) flag set if file is already random access (disables automatic SeekTest).
+2) Flag set if file is already random access (disables automatic SeekTest).
 
 =item Returns:
@@ -216 +216 @@
 =back
 
+=head1 OPTIONS
+
+=over 4
+
+=item B<NoBuffer>
+
+Avoid buffering sequential files.
+
+    $raf->{NoBuffer} = 1;
+
+When this option is set, old data is purged from the internal buffer before
+a read operation on a sequential file.  In this mode, memory requirements
+may be significantly reduced when reading sequential files, but seeking
+backward is limited to within the size of the internal buffer (which will be
+at least as large as the last returned data block), and seeking relative to
+the end of file is not allowed.
+
+=back
+
 =head1 AUTHOR
 
-Copyright 2003-2011 Phil Harvey (phil at owl.phy.queensu.ca)
+Copyright 2003-2021 Phil Harvey (philharvey66 at gmail.com)
 
 This library is free software; you can redistribute it and/or modify it