You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@pdfbox.apache.org by Maruan Sahyoun <sa...@fileaffairs.de> on 2014/01/29 10:44:52 UTC
[DISCUSS] GSoC Participation
Hi
shall we try to participate at GSoC? Needs a mentor though.
BR
Maruan Sahyoun
Re: [DISCUSS] GSoC Participation
Posted by Pasan Buddhika <pa...@gmail.com>.
Hi,
Good idea. Let's try
On Wed, Jan 29, 2014 at 3:14 PM, Maruan Sahyoun <sa...@fileaffairs.de>wrote:
> Hi
>
> shall we try to participate at GSoC? Needs a mentor though.
>
> BR
>
> Maruan Sahyoun
>
>
--
*W.A.Pasan Buddhika*
Contacts:
Mobile : 071 366 13 58
pasanbuddhika pasanbuddhika
Stay connect with me : <http://www.facebook.com/pasanbuddhika>
<http://lk.linkedin.com/in/pasanbuddhika>
<http://twitter.com/#!/pasanbuddhika>
My blog : www.pasanlive.blogspot.com
Re: [DISCUSS] GSoC Participation
Posted by Maruan Sahyoun <sa...@fileaffairs.de>.
There were several ideas floating around wo a real consensus. From my perspective PDFbox would benefit most if missing pieces could be implemented:
- shading types as Tilman suggested
- signature algorithms
- support for different character sets during PDF generation
- PDF optimization e.g. remove duplicate resources when merging PDFs or no longer needed ones during splitting a PDF
- PoC work could also be feasible of how to implement different PDF levels and standards in a similar, extendable manner.
I’d rather see us completing PDF core features than adding new functionality like table recognition, high level PDF creation API or OCR interface although these would be very beneficial functionalities.
In addition working on the documentation might be something although not for a ‚core‘ developer.
One question which needs to be answered is who would act as a mentor?
BR
Maruan Sahyoun
Am 13.02.2014 um 09:19 schrieb Andreas Lehmkühler <an...@lehmi.de>:
> Hi,
>
> for those who are still interested in GSoC, [1] has some information on how
> to participate. According to the mail it's maybe to late, but I would give
> it a try. I've shared a private link, only available to PMC-members, as some
> of the imformation seems to be private.
>
> BR
> Andreas Lehmkühler
>
> [1]
> https://mail-search.apache.org/pmc/private-arch/pdfbox-private/201401.mbox/%3C22705BFE-BE29-492A-BE12-0749317FBB75@apache.org%3E
>
>
>> John Hewson <jo...@jahewson.com> hat am 11. Februar 2014 um 23:52 geschrieben:
>>
>>
>> The ideas are supposed to be starting points for students to make their own
>> proposal, so give them some ideas for expanding/reducing the scope and they
>> can choose themselves.
>>
>> -- John
>>
>> On 11 Feb 2014, at 13:47, Tilman Hausherr <TH...@t-online.de> wrote:
>>
>>> Its unclear what the "size" of a participation must be. What I'd like to
>>> have is someone to implement shading types 6 and 7, and I think it would be
>>> 1-2 weeks of work. This would be perfect for a math student, or a computer
>>> science student who is specializing in graphics. My own math is from school
>>> 30 years ago and we never did Bézier curves, tensor-products and Bernstein
>>> polynomials so I can't do it without learning the math first.
>>>
>>> Tilman
>>
Re: [DISCUSS] GSoC Participation
Posted by Andreas Lehmkühler <an...@lehmi.de>.
Hi,
for those who are still interested in GSoC, [1] has some information on how
to participate. According to the mail it's maybe to late, but I would give
it a try. I've shared a private link, only available to PMC-members, as some
of the imformation seems to be private.
BR
Andreas Lehmkühler
[1]
https://mail-search.apache.org/pmc/private-arch/pdfbox-private/201401.mbox/%3C22705BFE-BE29-492A-BE12-0749317FBB75@apache.org%3E
> John Hewson <jo...@jahewson.com> hat am 11. Februar 2014 um 23:52 geschrieben:
>
>
> The ideas are supposed to be starting points for students to make their own
> proposal, so give them some ideas for expanding/reducing the scope and they
> can choose themselves.
>
> -- John
>
> On 11 Feb 2014, at 13:47, Tilman Hausherr <TH...@t-online.de> wrote:
>
> > Its unclear what the "size" of a participation must be. What I'd like to
> > have is someone to implement shading types 6 and 7, and I think it would be
> > 1-2 weeks of work. This would be perfect for a math student, or a computer
> > science student who is specializing in graphics. My own math is from school
> > 30 years ago and we never did Bézier curves, tensor-products and Bernstein
> > polynomials so I can't do it without learning the math first.
> >
> > Tilman
>
Re: [DISCUSS] GSoC Participation
Posted by John Hewson <jo...@jahewson.com>.
The ideas are supposed to be starting points for students to make their own proposal, so give them some ideas for expanding/reducing the scope and they can choose themselves.
-- John
On 11 Feb 2014, at 13:47, Tilman Hausherr <TH...@t-online.de> wrote:
> Its unclear what the "size" of a participation must be. What I'd like to have is someone to implement shading types 6 and 7, and I think it would be 1-2 weeks of work. This would be perfect for a math student, or a computer science student who is specializing in graphics. My own math is from school 30 years ago and we never did Bézier curves, tensor-products and Bernstein polynomials so I can't do it without learning the math first.
>
> Tilman
Re: [DISCUSS] GSoC Participation
Posted by Tilman Hausherr <TH...@t-online.de>.
Its unclear what the "size" of a participation must be. What I'd like to
have is someone to implement shading types 6 and 7, and I think it would
be 1-2 weeks of work. This would be perfect for a math student, or a
computer science student who is specializing in graphics. My own math is
from school 30 years ago and we never did Bézier curves, tensor-products
and Bernstein polynomials so I can't do it without learning the math first.
Tilman
Re: [DISCUSS] GSoC Participation
Posted by John Hewson <jo...@jahewson.com>.
I think the idea needs to be considerably more exciting to attract students - nobody want’s to fix the bugs that even we don’t want to fix!
There are some interesting users of PDFBox, see http://pdfliberation.wordpress.com/ for some possible ideas… lots of people using OCR there too.
> PDFBOX-1594 Add support for AES256 Encryption
Seems like a reasonable project.
-- John
On 29 Jan 2014, at 17:28, Fred Hansen <zw...@yahoo.com> wrote:
>
> IMHO a task for GSoC should be non-critical, localized, and not a user interface. A "non-critical" is one where PDFBOX development can continue without relying on the project result. A "localized" project is one that can be incorporated into the code base with few changes to the base. This will limit the effort required to learn about the system into which the effort will fit. A "user-interface" implements an interactive window or an API. I have low expectations of the capabilities of students for doing good designs in these areas.
>
> So I looked through JIRA for open projects meeting the above. Since I am not all that familiar with PDFBOX, some of my suggestions may be laughable and surely I have missed some. Nonetheless, here's what I found:
>
>
> PDFBOX-553 writing pdf file in Japanese, garbled
> PDFBOX-570 Windings font recognition + spacing issue
> PDFBOX-605 Better support for Type0 fonts
> PDFBOX-678 Support missing Text Rendering Modes when rendering a PDF
> PDFBOX-870 PDF-To-IMAGE output is not anti-aliased
> PDFBOX-1094 Pattern colorspace support
> PDFBOX-1594 Add support for AES256 Encryption
> (see also PDFBOX-1450 document how to encrypt with AES 256 )
> PDFBOX-1734 ImageIoUtil.WriteImage doesn't work with tiff images
> PDFBOX-1843 Find a way to test PDFToImage
>
>
>
>
>> ________________________________
>> From: John Hewson <jo...@jahewson.com>
>> To: "dev@pdfbox.apache.org" <de...@pdfbox.apache.org>
>> Sent: Wednesday, January 29, 2014 6:38 PM
>> Subject: Re: [DISCUSS] GSoC Participation
>>
>>
>>> - an idea which came up some years ago, was to implement a gui-interface to
>>> bundle some/all/future tools/features of pdfbox, like printing, rendering,
>>> preflight, split, merge etc.
>>
>> The AWT/Swing PDF viewer could do with rewriting. But does anyone want that? Maybe support for JavaFX?
>>
>>> - a high-level api to create pdfs
>>
>> I've been thinking about this recently and have come to the conclusion that it's really hard to do well.
>>
>>> - an advanced text extractor with table/column support
>>
>> The table stuff sounds a lot like Tabula? Do we really not have column support? We need that!
>>
>> I'll throw in some ideas too:
>>
>> - an interface for OCR engines to plug into the text extraction API. It could provide access to extracted images or allow badly encoded fonts to be passed to OCR one character or text run at a time.
>>
>> -
>>
>> -- John
>>
>>
>>> On 29 Jan 2014, at 03:20, Andreas Lehmkühler <an...@lehmi.de> wrote:
>>>
>>> Hi,
>>>
>>>> Maruan Sahyoun <sa...@fileaffairs.de> hat am 29. Januar 2014 um 10:44
>>>> geschrieben:
>>>>
>>>>
>>>> Hi
>>>>
>>>> shall we try to participate at GSoC? Needs a mentor though.
>>> That idea already came up from time to time and it didn't work for different
>>> reasons.
>>>
>>> So, to participate we need a mentor and or course at least one good idea to pe
>>> proposed.
>>>
>>> I won't act as mentor for different reasons but I'll try to help in the normal
>>> manner.
>>>
>>> IMO an appropriate idea shall not deal with pdf-specific low-level features,
>>> like linearization support, as I doubt that any possible student is familiar
>>> with the pdf-spec.
>>>
>>> So possible ideas could be:
>>>
>>> - an idea which came up some years ago, was to implement a gui-interface to
>>> bundle some/all/future tools/features of pdfbox, like printing, rendering,
>>> preflight, split, merge etc.
>>> - a high-level api to create pdfs
>>> - an advanced text extractor with table/column support
>>>
>>>
>>>> BR
>>>>
>>>> Maruan Sahyoun
>>>
>>> BR
>>> Andreas Lehmkühler
>>
Re: [DISCUSS] GSoC Participation
Posted by Fred Hansen <zw...@yahoo.com>.
IMHO a task for GSoC should be non-critical, localized, and not a user interface. A "non-critical" is one where PDFBOX development can continue without relying on the project result. A "localized" project is one that can be incorporated into the code base with few changes to the base. This will limit the effort required to learn about the system into which the effort will fit. A "user-interface" implements an interactive window or an API. I have low expectations of the capabilities of students for doing good designs in these areas.
So I looked through JIRA for open projects meeting the above. Since I am not all that familiar with PDFBOX, some of my suggestions may be laughable and surely I have missed some. Nonetheless, here's what I found:
PDFBOX-553 writing pdf file in Japanese, garbled
PDFBOX-570 Windings font recognition + spacing issue
PDFBOX-605 Better support for Type0 fonts
PDFBOX-678 Support missing Text Rendering Modes when rendering a PDF
PDFBOX-870 PDF-To-IMAGE output is not anti-aliased
PDFBOX-1094 Pattern colorspace support
PDFBOX-1594 Add support for AES256 Encryption
(see also PDFBOX-1450 document how to encrypt with AES 256 )
PDFBOX-1734 ImageIoUtil.WriteImage doesn't work with tiff images
PDFBOX-1843 Find a way to test PDFToImage
>________________________________
> From: John Hewson <jo...@jahewson.com>
>To: "dev@pdfbox.apache.org" <de...@pdfbox.apache.org>
>Sent: Wednesday, January 29, 2014 6:38 PM
>Subject: Re: [DISCUSS] GSoC Participation
>
>
>> - an idea which came up some years ago, was to implement a gui-interface to
>> bundle some/all/future tools/features of pdfbox, like printing, rendering,
>> preflight, split, merge etc.
>
>The AWT/Swing PDF viewer could do with rewriting. But does anyone want that? Maybe support for JavaFX?
>
>> - a high-level api to create pdfs
>
>I've been thinking about this recently and have come to the conclusion that it's really hard to do well.
>
>> - an advanced text extractor with table/column support
>
>The table stuff sounds a lot like Tabula? Do we really not have column support? We need that!
>
>I'll throw in some ideas too:
>
>- an interface for OCR engines to plug into the text extraction API. It could provide access to extracted images or allow badly encoded fonts to be passed to OCR one character or text run at a time.
>
>-
>
>-- John
>
>
>> On 29 Jan 2014, at 03:20, Andreas Lehmkühler <an...@lehmi.de> wrote:
>>
>> Hi,
>>
>>> Maruan Sahyoun <sa...@fileaffairs.de> hat am 29. Januar 2014 um 10:44
>>> geschrieben:
>>>
>>>
>>> Hi
>>>
>>> shall we try to participate at GSoC? Needs a mentor though.
>> That idea already came up from time to time and it didn't work for different
>> reasons.
>>
>> So, to participate we need a mentor and or course at least one good idea to pe
>> proposed.
>>
>> I won't act as mentor for different reasons but I'll try to help in the normal
>> manner.
>>
>> IMO an appropriate idea shall not deal with pdf-specific low-level features,
>> like linearization support, as I doubt that any possible student is familiar
>> with the pdf-spec.
>>
>> So possible ideas could be:
>>
>> - an idea which came up some years ago, was to implement a gui-interface to
>> bundle some/all/future tools/features of pdfbox, like printing, rendering,
>> preflight, split, merge etc.
>> - a high-level api to create pdfs
>> - an advanced text extractor with table/column support
>>
>>
>>> BR
>>>
>>> Maruan Sahyoun
>>
>> BR
>> Andreas Lehmkühler
>
>
Re: [DISCUSS] GSoC Participation
Posted by John Hewson <jo...@jahewson.com>.
> - an idea which came up some years ago, was to implement a gui-interface to
> bundle some/all/future tools/features of pdfbox, like printing, rendering,
> preflight, split, merge etc.
The AWT/Swing PDF viewer could do with rewriting. But does anyone want that? Maybe support for JavaFX?
> - a high-level api to create pdfs
I've been thinking about this recently and have come to the conclusion that it's really hard to do well.
> - an advanced text extractor with table/column support
The table stuff sounds a lot like Tabula? Do we really not have column support? We need that!
I'll throw in some ideas too:
- an interface for OCR engines to plug into the text extraction API. It could provide access to extracted images or allow badly encoded fonts to be passed to OCR one character or text run at a time.
-
-- John
> On 29 Jan 2014, at 03:20, Andreas Lehmkühler <an...@lehmi.de> wrote:
>
> Hi,
>
>> Maruan Sahyoun <sa...@fileaffairs.de> hat am 29. Januar 2014 um 10:44
>> geschrieben:
>>
>>
>> Hi
>>
>> shall we try to participate at GSoC? Needs a mentor though.
> That idea already came up from time to time and it didn't work for different
> reasons.
>
> So, to participate we need a mentor and or course at least one good idea to pe
> proposed.
>
> I won't act as mentor for different reasons but I'll try to help in the normal
> manner.
>
> IMO an appropriate idea shall not deal with pdf-specific low-level features,
> like linearization support, as I doubt that any possible student is familiar
> with the pdf-spec.
>
> So possible ideas could be:
>
> - an idea which came up some years ago, was to implement a gui-interface to
> bundle some/all/future tools/features of pdfbox, like printing, rendering,
> preflight, split, merge etc.
> - a high-level api to create pdfs
> - an advanced text extractor with table/column support
>
>
>> BR
>>
>> Maruan Sahyoun
>
> BR
> Andreas Lehmkühler
Re: [DISCUSS] GSoC Participation
Posted by Andreas Lehmkühler <an...@lehmi.de>.
Hi,
> Maruan Sahyoun <sa...@fileaffairs.de> hat am 29. Januar 2014 um 10:44
> geschrieben:
>
>
> Hi
>
> shall we try to participate at GSoC? Needs a mentor though.
That idea already came up from time to time and it didn't work for different
reasons.
So, to participate we need a mentor and or course at least one good idea to pe
proposed.
I won't act as mentor for different reasons but I'll try to help in the normal
manner.
IMO an appropriate idea shall not deal with pdf-specific low-level features,
like linearization support, as I doubt that any possible student is familiar
with the pdf-spec.
So possible ideas could be:
- an idea which came up some years ago, was to implement a gui-interface to
bundle some/all/future tools/features of pdfbox, like printing, rendering,
preflight, split, merge etc.
- a high-level api to create pdfs
- an advanced text extractor with table/column support
> BR
>
> Maruan Sahyoun
BR
Andreas Lehmkühler