Adding ability to extract text to an io.Reader

[robarchibald]

Thanks for the awesome library! I thought it would be useful to have the ability to extract text from a PDF file easily so I added it to pdflib. The charmap capability was a beast, but I got it working for the files I tested on. Thanks for considering this pull request

[coveralls]

Coverage decreased (-0.6%) to 50.672% when pulling 99d8987 on EndFirstCorp:master into c0da03d on hhrutter:master.

[hhrutter]

Thx for your PR!
Appreciated 👍
Sorry I did not see it until I committed my squashed master branch :(
I'll get back to you.

[hhrutter]

Please share your specific intention of returning text.
Do you need text for a specific page or the whole file?
Please provide some tests and (go)doc so I get the idea of your usecase.
The api interfaces with files and directories for the moment so we want to stay consistent.
If there is a usecase for providing extracted content via io.Reader maybe that's a missing layer (in between). Like I said I squashed the last couple of commits together including the one you forked off, could you be so kind and provide your changes based off 8f3ac5c. Thank you so much.

[robarchibald]

I appreciate you taking the time to look at this pull request. It turns out that since submitting the pull request I created my own library to convert PDF’s to text at https://github.com/EndFirstCorp/pdf2txt. My use case for this is that I’m filling out a pipeline from a web server file upload. On file upload, you get access to a multipart.File. Rather than saving that multipart.File to a standard OS file, my goal was to minimize the I/O and 1) read in the file and convert it to text and then 2) output the text file to disk. I wanted it to work with an io.Reader, not a io.ReaderAt like the typical implementation of PDF parsers are. So, that’s what I’ve done.

If you’re still interested in implementing this pull request, I can look at resubmitting.

[hhrutter]

Understood. If you can resubmit your text extraction code so that it is consistent with the existing extraction code for images, fonts and content and supply test code as well I am happy to merge in. Basically what that means is writing out to a file and not a reader. Thank you.

[hhrutter]

Text extraction will definitely be an additional functionality at some point but
since this PR originated from an old version for now I am closing it.