Package: pdfminer-data
Version: 20221105+dfsg-1
Severity: normal
Tags: patch
Dear Maintainer,
Some data files in the package are corrupted.
$ for f in /usr/share/pdfminer/*.gz; do gunzip -t "$f"; done
gzip: /usr/share/pdfminer/to-unicode-Adobe-CNS1.pickle.gz: unexpected end of
file
gzip: /usr/share/pdfminer/to-unicode-Adobe-GB1.pickle.gz: unexpected end of file
gzip: /usr/share/pdfminer/to-unicode-Adobe-Japan1.pickle.gz: unexpected end of
file
gzip: /usr/share/pdfminer/to-unicode-Adobe-Korea1.pickle.gz: unexpected end of
file
This is caused by not closing the gzip files in conv_cmap.py. After applying
0002-Avoid-timestamps-in-gzip-compressed-file-and-use-com.patch, the following
patch fixes the bug.
--- tools/conv_cmap.py.orig 2024-09-28 03:36:58.077510467 +0900
+++ tools/conv_cmap.py 2024-09-28 03:37:33.928126725 +0900
@@ -202,6 +202,7 @@
# 2.7
fgz = gzip.GzipFile('', 'wb', 9, fp, 0.)
converter.dump_unicodemap(fgz)
+ fgz.close()
return
