Thomas Mortagne created TIKA-4296:
-------------------------------------
Summary: "Parameter must be 1-based, but is -1" when using Tika
with PDFBox 2.0.32
Key: TIKA-4296
URL: https://issues.apache.org/jira/browse/TIKA-4296
Project: Tika
Issue Type: Bug
Components: parser
Affects Versions: 2.9.2
Reporter: Thomas Mortagne
Attachments: pdf.pdf
I just upgraded my pdfbox dependency to 2.0.32 and any Tika#parseToString of a
pdf file seems to produce the following warning:
{noformat}
WARN o.apache.pdfbox.text.PDFTextStripper - Parameter must be 1-based, but is
-1
{noformat}
The behavior is the same as with 2.0.31, it's just that pdfbox is apparently
not too happy anyore with the way it's used by Tika.
Just in case it's not actually any file, here is one with which I reproduce:
[^pdf.pdf]
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
