Plain Text Version of Rtf Scan

boomer3200

It would be helpful if Zotero expanded and applied styles to keyed references in text files. Essentially, I would like to use a plain text version of the current "Rtf Scan" feature to work with pandoc (extended markdown) and troff source documents, and the program would potentially gain further users among the Scrivener crowd on OS X as well as other Unix-based systems. Perhaps the output could optionally include underscores as markup for emphasis around book and journal titles (e.g. _Title_).

Dan Plassche

--
boomer3200

anders_royce

This has been suggested before. Here is a recent thread: http://forums.zotero.org/discussion/12176/plaintext-scanning/

boomer3200

Okay, well are isolated negative responses to the post you mention and a previous one I encountered authoritative?[1]

Consider the issue. Omitting the rtf code from an existing parser and output would not be prohibitively difficult or time consuming. Rich text files, which may follow any of several different version specifications, have not been implemented for standardized display across applications and operating systems. Those opposing a plain text option in both discussions are short-sighted to respond that one can simply use LaTeX/BibTeX instead.[2] Many people find the barrier of entry in learning to use (La)Tex too high to bother. I have used Tex, but find the command codes cumbersome to enter and non-human readable in contrast to pandoc, while the ouput is less open to fine-grained customization and positioning of text than troff.

[1]: http://forums.zotero.org/discussion/3591/integration-with-word-processors-via-rtf-scan/

[2]: Especially in http://forums.zotero.org/discussion/12176/plaintext-scanning/

Regards,

Dan Plassche


--
boomer3200

adamsmith

don't misunderstand my comment on the other thread- I don't "oppose" plain text scanning (nor do I think anyone else does)- I just think that it's not worth investing much time in a half-baked solution and I think doing this well is a pretty substantial effort - one that - considering that Zotero works well with the three by far most common writing products among academics - shouldn't be a priority for the core developers.

I think the RTF scan is used by so few people and with relatively poor results that I don't believe it'd help much extending the current version. But for anyone who wants to - rtfScan.js in the main folder of the Zotero.jar

And as I say on the other thread - there is always the option to write an add-on or contribute code directly if you feel soffuciently strongly about this.

boomer3200

Thanks for your reply, adamsmith. In brief I understand your priorities of the moment, but would also respond that design and usage levels for at least two of those applications (Word and OpenOffice) are volatile, while dragging citations to the third (LaTeX) is time consuming in contrast to alternatives, unlike auto-formatting a plain text file.

Word's recent changes in the 2007 edition caused problems for integrating Zotero on top of a sometimes unreliable (docx corruption) and harder to use (ribbon) application for prior advanced users, which are some of the reasons that I became increasingly concerned about the direction of the program. Concentrated (funded) development of OpenOffice in the future also appears uncertain, and although the suite will remain available, I have not seen the program commonly used by scholars due to the academic licensing of Word.

As far as coding a plain text processor, I would first develop a fast and portable external solution for personal use outside of Zotero before going through the javascript. Zotero still serves as an excellent tool to collect and export a bibliographic database for further processing.

Dan

--
boomer3200