Re: Xref parsing performance

Am 28.02.2015 um 19:54 schrieb Andreas Lehmkuehler: Am 28.02.2015 um 18:34 schrieb Maruan Sahyoun: Am 28.02.2015 um 18:18 schrieb Andreas Lehmkuehler : Am 28.02.2015 um 18:07 schrieb Maruan Sahyoun: Hi, Am 28.02.2015 um 17:53 schrieb Andreas Lehmkuehler : Am 28.02.2015 um 17:49 schrieb Ma

Re: Xref parsing performance

Am 28.02.2015 um 18:34 schrieb Maruan Sahyoun: Am 28.02.2015 um 18:18 schrieb Andreas Lehmkuehler : Am 28.02.2015 um 18:07 schrieb Maruan Sahyoun: Hi, Am 28.02.2015 um 17:53 schrieb Andreas Lehmkuehler : Am 28.02.2015 um 17:49 schrieb Maruan Sahyoun: Hi, Am 28.02.2015 um 17:32 schrieb An

Re: Xref parsing performance

Am 28.02.2015 um 18:18 schrieb Andreas Lehmkuehler : > Am 28.02.2015 um 18:07 schrieb Maruan Sahyoun: >> Hi, >> >> Am 28.02.2015 um 17:53 schrieb Andreas Lehmkuehler : >> >>> Am 28.02.2015 um 17:49 schrieb Maruan Sahyoun: Hi, Am 28.02.2015 um 17:32 schrieb Andreas Lehmkuehler :

Re: Xref parsing performance

Am 28.02.2015 um 18:07 schrieb Maruan Sahyoun: Hi, Am 28.02.2015 um 17:53 schrieb Andreas Lehmkuehler : Am 28.02.2015 um 17:49 schrieb Maruan Sahyoun: Hi, Am 28.02.2015 um 17:32 schrieb Andreas Lehmkuehler : Hi Am 28.02.2015 um 16:47 schrieb Tilman Hausherr: Hi Andrea, While a speed imp

Re: Xref parsing performance

Hi, Am 28.02.2015 um 17:53 schrieb Andreas Lehmkuehler : > Am 28.02.2015 um 17:49 schrieb Maruan Sahyoun: >> Hi, >> >> Am 28.02.2015 um 17:32 schrieb Andreas Lehmkuehler : >> >>> Hi >>> >>> Am 28.02.2015 um 16:47 schrieb Tilman Hausherr: Hi Andrea, While a speed improvement in

Re: Xref parsing performance

Am 28.02.2015 um 17:58 schrieb Tilman Hausherr: Am 28.02.2015 um 17:49 schrieb Maruan Sahyoun: >> >>- why change only one of the members of that cosobjectkey class to int? >>According to the spec, both are integers. Maybe there's a good reason, but I'd >>like to know. >ASFAIK there is no good r

Re: Xref parsing performance

Am 28.02.2015 um 17:49 schrieb Maruan Sahyoun: Hi, Am 28.02.2015 um 17:32 schrieb Andreas Lehmkuehler : Hi Am 28.02.2015 um 16:47 schrieb Tilman Hausherr: Hi Andrea, While a speed improvement in parsing of large files would be much appreciated (especially by the TIKA users), there are sever

Re: Xref parsing performance

Am 28.02.2015 um 17:49 schrieb Maruan Sahyoun: >> >>- why change only one of the members of that cosobjectkey class to int? >>According to the spec, both are integers. Maybe there's a good reason, but I'd >>like to know. >ASFAIK there is no good reason not to change both to int. as the offset i

Re: Xref parsing performance

mmm... are you using the tip of the "xref" branch? Because it shouldn't use any jdk7 stuff and it compiles and runs fine on my machine. I'm using Ubuntu and jdk1.6.0_45 and I have: [INFO] BUILD SUCCESS I changed the generation number to int because in the xref table it's a 5 digit number so it fit

Re: Xref parsing performance

Hi, Am 28.02.2015 um 17:32 schrieb Andreas Lehmkuehler : > Hi > > Am 28.02.2015 um 16:47 schrieb Tilman Hausherr: >> Hi Andrea, >> >> While a speed improvement in parsing of large files would be much appreciated >> (especially by the TIKA users), there are several problems with your change: > +

Re: Xref parsing performance

Hi Am 28.02.2015 um 16:47 schrieb Tilman Hausherr: Hi Andrea, While a speed improvement in parsing of large files would be much appreciated (especially by the TIKA users), there are several problems with your change: +1 - don't do changes that need JDK7 or higher even if they are cool. We us

Re: Xref parsing performance

Hi Andrea, While a speed improvement in parsing of large files would be much appreciated (especially by the TIKA users), there are several problems with your change: - don't do changes that need JDK7 or higher even if they are cool. We use JDK6 currently. - regressions: Error converting f

Re: Xref parsing performance

looked at it quickly - very nice! Maruan Am 27.02.2015 um 16:34 schrieb Andrea Vacondio : > Hi, > few days ago I was profiling PDFBox when loading medium/large size > documents and I think I found something. > If you try loading the document > http://www.adobe.com/devnet/acrobat/pdfs/pdf_refere

Xref parsing performance

Hi, few days ago I was profiling PDFBox when loading medium/large size documents and I think I found something. If you try loading the document http://www.adobe.com/devnet/acrobat/pdfs/pdf_reference_1-7.pdf you'll see it takes quite some time and that's mostly spent in the XrefTrailerResolver.getCo