#945 PDF file corrupted (file size 0) after writing XMP data

[Anonymous]

PDF files are corrupted if the abstract contains single quotes. There may be other characters triggering this bug as well, but so far I only found examples where removing single quotes from the abstract solved this issue. The PDF files have file size zero after writing the XMP data to the file.

[Henrik Seidel]

Hello,

I was the original submitter of this bug. Meanwhile, it happens more and more often that I experience this issue. Here is an example of a PDF document that demonstrates the bug:

@ARTICLE{Oshikawa2002,
author = {K. Oshikawa and K. Yanagisawa and S. Tominaga and Y. Sugiyama},
title = {Expression and function of the ST2 gene in a murine model of allergic
airway inflammation.},
journal = {Clin Exp Allergy},
year = {2002},
volume = {32},
pages = {1520--1526},
number = {10},
month = {Oct},
abstract = {We have recently reported that soluble ST2 protein levels are elevated
in the sera of patients with asthma, and correlate well with the
severity of asthma exacerbation. However, the role, function, and
kinetics of soluble ST2 expression in asthma remain unclear.The objective
of the present study was to clarify the function and kinetics of
soluble murine (m) ST2 expression in a murine asthma model.We analyzed
the kinetics of gene and protein expression of mST2 in sera or lung
tissue after allergen (ovalbumin; OVA) challenge in a murine model
of allergic airway inflammation, the effects of mST2 protein on OVA-induced
Th2 cytokine production in vitro from splenocytes of sensitized mice,
and the effects of soluble mST2 on Th2-dependent allergic airway
inflammation by in vivo gene transfer of mST2.Serum mST2 protein
levels increased to the maximal level 3 h after the allergen challenge,
before serum IL-5 levels peaked. The mRNA expression of mST2 in lung
tissue was induced after the allergen challenge, while that in the
spleen was constitutively detected. Furthermore, pre-treatment with
mST2 protein significantly inhibited the production of IL-4 and IL-5,
but not IFN-gamma, from OVA-stimulated splenocytes in vitro, and
intravenous mST2 gene transfer resulted in a drastic reduction in
the number of eosinophils and in the levels of IL-4 and IL-5 in bronchoalveolar
lavage fluid, compared with those in response to transfer of non-coding
plasmid vector or of lipid alone.These results suggest that increases
in endogenous mST2 protein after allergen exposure may modulate Th2-mediated
airway inflammation, and that in vivo gene transfer of mST2 can be
applicable to use in a novel immunotherapy for allergic diseases.},
file = {Oshikawa2002.pdf:Endometriosis\\Oshikawa2002.pdf:PDF},
institution = {Division of Pulmonary Medicine, Department of Medicine, Jichi Medical
School, Tochigi, Japan. kokyu2@jichi.ac.jp},
keywords = {Allergens, diagnostic use; Analysis of Variance; Animals; Bronchoalveolar
Lavage Fluid, immunology; Female; Gene Expression; Genetic Engineering;
Genetic Vectors, administration /&/ dosage; Interferon-gamma, blood;
Interleukin-4, blood; Interleukin-5, blood; Lung, metabolism; Membrane
Proteins, blood/genetics/metabolism; Mice; Mice, Inbred BALB C; Models,
Animal; Ovalbumin, diagnostic use; Receptors, Interleukin; Respiratory
Hypersensitivity, metabolism},
language = {eng},
medline-pst = {ppublish},
owner = {sgpdc},
pii = {1494},
pmid = {12372135},
timestamp = {2011.03.09},
url = {http://onlinelibrary.wiley.com/doi/10.1046/j.1365-2745.2002.01494.x/abstract}
}

[nsubiron]

PDFs corrupted after writing XMP data to them, size is not zero but sometimes parts of the pdf become unreadable (and you don't see till you get to that point). Acrobat Reader gives error 110, Sumatra just displays the pdf with missing text.

Here an example of file that gets corrupted http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.53.7505 (attaching metadata from google scholar search) using JabRef version 2.10.

I've lost hours organizing my library now to see that files might or might not be corrupted. This is a critical bug and IMO should be warned in your homepage.

WRITING XMP METADATA MIGHT CORRUPT THE PDF.

-- angry user leaves slamming the door --

[Oliver Kopp]

Hendrik, could you please mail me the PDF?