[ https://issues.apache.org/jira/browse/PDFBOX-4371?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16703689#comment-16703689 ]
Tilman Hausherr commented on PDFBOX-4371: ----------------------------------------- I did find a few examples where the rotation magic produces a poor result, e.g. commoncrawl3/FR/FRQYA5JFUK6ZONSKW6A24P2TWJMGXDA6 This is because there is a slight rotation difference for every glyph. This would count as one line in classic extract, but as different rotations in the rotationMagic extract. > Improve ExtractText utility so that it can extract rotated text automatically > ----------------------------------------------------------------------------- > > Key: PDFBOX-4371 > URL: https://issues.apache.org/jira/browse/PDFBOX-4371 > Project: PDFBox > Issue Type: Improvement > Components: Text extraction, Utilities > Affects Versions: 2.0.12, 3.0.0 PDFBox > Reporter: Tilman Hausherr > Assignee: Tilman Hausherr > Priority: Major > Fix For: 2.0.13, 3.0.0 PDFBox > > Attachments: ExtractAngledText.java > > > In a first step, detect all rotations by analyzing the effective text > rendering matrix. In a second step, do a text extraction for each rotation by > prepending an appropriate transform to the page content stream (so that our > text has angle == 0) and then filtering any rotated text. Test file: the file > from PDFBOX-4368. -- This message was sent by Atlassian JIRA (v7.6.3#76005) --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@pdfbox.apache.org For additional commands, e-mail: dev-h...@pdfbox.apache.org