- 1 Personal Information
- 2 Project Information
- 2.1 Project Title
- 2.2 Brief project summary
- 2.3 Detailed project description
- 2.4 Deliverables
- 2.5 Development Schedule
- 2.6 Time availability
- 2.7 Past Experiences
- 2.8 Why me
- 2.9 References
Name: Richa Jain
IRC Nick: richa (#brlcad on freenode)
Synchronize wiki with Docbook
Brief project summary
BRL-CAD’s documentation is available in a variety of formats but the best way of maintaining documentation is docbook XML. However, as we know that a general user finds it easier to edit a wiki page rather than a docbook. Conversion between XML markup (docbook) and wiki syntax is not that easy. Using tools like pandoc for this, some markups are lost and ultimately the data is ruined. So, the main aim of this project is to Synchronize wiki and docbook so that any changes made to the wiki appears in the docbook and vice-versa without any loss of data. The project will be implemented as a separate mediawiki extension which shows the docbook XML in the edit window and after they have edited it, the XML is processed to produce the HTML web output. If a user prefers to edit wiki markup, then after the editing, the docbook XML can be downloaded.
Detailed project description
Why is Docbook format important?
Docbook is an XML language. Docbook provides you a large number of semantic element tags. These tags give some structure to the document and this structured document can be published in a variety of formats including HTMl, XHTML, PDF, etc. without any change to the source code. Most of the organisations maintain their documentation in the docbook format due to its easy maintainance. BRL-CAD also focuses on maintaining its documentation in docbook format so that it can be more easily managed by revision control system.
Tools available for conversion between Wiki to Docbook and Vice-Versa:
There are some tools available which convert docbook to wiki or vice-versa which can be used for this project:
- Pandoc converts from one markup to another. However, while conversion between wiki and docbook, some of the tags are lost and hence the data is not preserved.
- Mediawiki’s Collection Extension allows user to collect articles and generate downloadable versions in different formats (including XML format too). It is configured by default to use the server http://tools.pediapress.com/ . For high traffic, we need to set up mwlib and setup our own render server which will generate the docbook XML output.
- Fedora’s Beacon project which was a done as GSoC 2009 is an XML editor which allows user to upload, create or edit docbook and parse the XML and return the HTML weboutput.
Basic Layout of the project:
- User will be given two edit options - wiki markup or Docbook XML.
- If a user chooses editing Docbook, the docbook XML will be displayed in the editing window rather than wiki syntax. When the user clicks on “save”, the entire XML is processed and converted to wikitext and the HTML output is displayed.
- If a user prefers to edit wiki markup, then after clicking “save” button the wiki markup is converted to docbook and the user is given an option to download the XML file.
Possible approaches for the project
There are basically two approaches for this project:
1. We can use a combination of Mediawiki's Collection Extension and Beacon Editor inorder to fufill the goal of this project.
Conversion from wiki to Docbook - Mediawiki's Collection Extension focuses on creating books by collecting articles and exporting this book to a variety of formats including the docbook XML format. This extension works as a special page. For our project, instead of this extension being implemented as a special page, we need to call this extension whenever a user chooses to edit the wiki.
Conversion from Docbook to wiki - Since Beacon project is an XML editor. This can be used to fetch XML to the edit window and then parsing this XML to generate the HTML web output.
2. Instead of using combination of different tools, we can define our own mediawiki extension which intelligently maps wiki markup to docbook XML markup and serves both the purposes simultaneously. This can be done with mediawiki parser functions.
I am strongly in favour of following the first approach since there are already tools available which can fulfill the purpose. This may need a few quick experiements to be done with these tools. I will utilize the community bonding period and figure out which of them would require lesser effort and is more efficient.
- A mediawiki extension having two edit options - Wiki and Docbook.
- Providing XML markup instead of wiki when a user chooses "Edit via Docbook".
- Mapping Docbook and wiki by using any of the approaches described above such that there exists a synchronization between the two edits.
- Adding an option to download the XML file once a user has finished editing wiki markup.
- Creating documentation which describes the technologies used for future developers ( Documentation for the code will be written as I write the code )
I will start working on my project as my exams will get over (on 30th April). The basic plan and estimated timeline of my project are: (timeline may change a little during the project)
- May 1 to May 18 (Community bonding period): Investigating the two approaches described above, making few experiments with Beacon and Collection Extension and deciding which of the two will be better.
- May 19 to May 31: Making a skeleton for the extension and giving two edit options - wiki and docbook.
- June 1 to July 15: Implementing the backend of the extension. This involves fetching of XML to the edit window and mapping XML to wiki and vice-versa using any of the two approaches.
- July 16 to July 31: Testing and bug fixing
- August 1 to August 10: Writing relevant Documentation
- August 11 to August 18: Final Evaluations
- Time Zone: UTC+5:30 (IST-India)
- Typical working hours: 12 noon to 1 am before July, 6pm to 3am after July
I am quite flexible with the timings and will try to dedicate most of my time to this project.
- My open source projects can be seen on github.
I will work hard and bring this project to completion. I am always logged in to IRC (#brlcad) whenever I work and also do respond to the emails quickly whenever I am awake. I will keep a repository of this project on github and make frequent commits. I will always be in touch with my mentor and the community and taking help from people around whenever I am stuck.