[ https://issues.apache.org/jira/browse/TIKA-2509?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16326160#comment-16326160 ]
Dave Meikle commented on TIKA-2509: ----------------------------------- Created new improvement ticket for the Python path configuration noted in this ticket > TesseractOCRParser ignores configured ImageMagickPath in processImage method > ---------------------------------------------------------------------------- > > Key: TIKA-2509 > URL: https://issues.apache.org/jira/browse/TIKA-2509 > Project: Tika > Issue Type: Bug > Components: ocr > Affects Versions: 1.16, 1.17 > Reporter: Richard Jones > Assignee: Dave Meikle > Priority: Major > Fix For: 1.18 > > > The TesseractOCRParser class uses the configured ImageMagickPath in method > hasImageMagick to determine whether ImageMagick is present. Ref: > String ImageMagick = config.getImageMagickPath() + getImageMagickProg(); > BUT then completely ignores the configured path in the processImage method > meaning ImageMagick has to be present on system path (so what's the point of > the ImageMagickPath config setting). > The doOCR method on the other hand DOES use the configured tesseractPath. > Incidentally I notice there is no equivalent PythonPath config setting even > though Python is attempted to be found/used. > Some consistency would be appreciated so that ImageMagick and Python don't > have to be present on the system path. i.e. follow the model already in > place for finding/using Tesseract. -- This message was sent by Atlassian JIRA (v7.6.3#76005)