[
https://issues.apache.org/jira/browse/TIKA-3364?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17330824#comment-17330824
]
David Pilato commented on TIKA-3364:
------------------------------------
So I trie this:
{code:java}
PDFParser pdfParser = new PDFParser();
DefaultParser defaultParser;
pdfParser.setExtractAnnotationText(false);
if (!fs.getOcr().isEnabled()) {
logger.debug("OCR is disabled. Even though it's detected, it
must be disabled explicitly");
defaultParser = new DefaultParser(
MediaTypeRegistry.getDefaultRegistry(),
new ServiceLoader(),
Collections.singletonList(TesseractOCRParser.class));
} else {
logger.debug("OCR is activated.");
if (ExternalParser.check("tesseract")) {
logger.debug("OCR strategy for PDF documents is [{}] and
tesseract was found.", fs.getOcr().getPdfStrategy());
pdfParser.setOcrStrategy(fs.getOcr().getPdfStrategy());
} else {
logger.debug("But Tesseract is not installed so we won't
run OCR.");
pdfParser.setOcrStrategy("no_ocr");
}
defaultParser = new DefaultParser(
MediaTypeRegistry.getDefaultRegistry(),
new ServiceLoader(),
Collections.singletonList(PDFParser.class));
}
parser = new AutoDetectParser(defaultParser, pdfParser);
{code}
And it seems to be producing the same effect. I'm probably missing something.
When I run it with this configuration, the extracted text is actually:
{code:txt}
\nDummy PDF file\n\nDummy PDF file\n\n\n\tDummy PDF file\n\n
{code}
So the text is extracted 3 times.
When I disable OCR with {{pdfParser.setOcrStrategy("no_ocr")}}, I'm getting:
{code:txt}
\nDummy PDF file\n\n\n\tDummy PDF file\n\n
{code}
> PDF Content is extracted twice
> ------------------------------
>
> Key: TIKA-3364
> URL: https://issues.apache.org/jira/browse/TIKA-3364
> Project: Tika
> Issue Type: Bug
> Components: parser
> Affects Versions: 1.26
> Reporter: David Pilato
> Priority: Major
> Attachments: Screenshot from 2021-04-23 10-15-22.png, issue-1097.pdf,
> tika-bookmarks-config.xml
>
>
> Hi
> Coming from [this issue in FSCrawler
> project|https://github.com/dadoonet/fscrawler/issues/1097], I can see that
> the text from the PDF document is extracted more than once although PDFBox
> seems to extract it only once.
> I attached the PDF.
> When I run:
> {code:sh}
> wget https://downloads.apache.org/pdfbox/2.0.23/pdfbox-app-2.0.23.jar
> java -jar pdfbox-app-2.0.23.jar ExtractText -console issue-1097.pdf
> {code}
> I'm getting:
> {code:sh}
> Dummy PDF file
> {code}
> But with Tika:
> {code:sh}
> wget https://downloads.apache.org/tika/tika-app-1.26.jar
> java -jar tika-app-1.26.jar
> {code}
> I'm getting:
> {code:xml}
> <?xml version="1.0" encoding="UTF-8"?><html
> xmlns="http://www.w3.org/1999/xhtml";>
> <head>
> <meta name="pdf:PDFVersion" content="1.4"/>
> <meta name="xmp:CreatorTool" content="Writer"/>
> <meta name="pdf:hasXFA" content="false"/>
> <meta name="access_permission:modify_annotations" content="true"/>
> <meta name="access_permission:can_print_degraded" content="true"/>
> <meta name="dc:creator" content="Evangelos Vlachogiannis"/>
> <meta name="dcterms:created" content="2007-02-23T15:56:37Z"/>
> <meta name="dc:format" content="application/pdf; version=1.4"/>
> <meta name="pdf:docinfo:creator_tool" content="Writer"/>
> <meta name="access_permission:fill_in_form" content="true"/>
> <meta name="pdf:encrypted" content="false"/>
> <meta name="Content-Length" content="13264"/>
> <meta name="X-TIKA:digest:MD5" content="2942bfabb3d05332b66eb128e0842cff"/>
> <meta name="pdf:hasMarkedContent" content="false"/>
> <meta name="Content-Type" content="application/pdf"/>
> <meta name="pdf:docinfo:creator" content="Evangelos Vlachogiannis"/>
> <meta name="X-Parsed-By" content="org.apache.tika.parser.DefaultParser"/>
> <meta name="X-Parsed-By" content="org.apache.tika.parser.pdf.PDFParser"/>
> <meta name="creator" content="Evangelos Vlachogiannis"/>
> <meta name="meta:author" content="Evangelos Vlachogiannis"/>
> <meta name="meta:creation-date" content="2007-02-23T15:56:37Z"/>
> <meta name="created" content="2007-02-23T15:56:37Z"/>
> <meta name="X-TIKA:digest:SHA256"
> content="3df79d34abbca99308e79cb94461c1893582604d68329a41fd4bec1885e6adb4"/>
> <meta name="access_permission:extract_for_accessibility" content="true"/>
> <meta name="access_permission:assemble_document" content="true"/>
> <meta name="xmpTPg:NPages" content="1"/>
> <meta name="Creation-Date" content="2007-02-23T15:56:37Z"/>
> <meta name="resourceName" content="issue-1097.pdf"/>
> <meta name="pdf:hasXMP" content="false"/>
> <meta name="access_permission:extract_content" content="true"/>
> <meta name="access_permission:can_print" content="true"/>
> <meta name="Author" content="Evangelos Vlachogiannis"/>
> <meta name="producer" content="OpenOffice.org 2.1"/>
> <meta name="access_permission:can_modify" content="true"/>
> <meta name="pdf:docinfo:producer" content="OpenOffice.org 2.1"/>
> <meta name="pdf:docinfo:created" content="2007-02-23T15:56:37Z"/>
> <title/>
> </head>
> <body><div class="page"><p/>
> <p>Dummy PDF file</p>
> <p/>
> </div>
> <ul> <li>Dummy PDF file</li>
> </ul>
> </body></html>
> {code}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
