GSoC 2009 Search and Replace by meet

This article is ARCHIVED. It is a part of the Scribus Community's history and is kept here for reference and as a representation of the contributed efforts. It is protected from editing, but its source can be copied.

Search and Replace

Rationale

Text is a substantial part of many Scribus documents. 'Search and Replace' is one of the most fundamental and widely used text manipulation operations. The current implementation lacks many features while having a non-modular UI.

Reimplementing it with new features will increase both the usability and popularity of Scribus.

Motivation

Scribus is the leading Open Source Desktop Publishing software. However, commercial programs appear to have better implementation of such features as 'Search and Replace'. Being widely used, this feature has to be improved. I have a lot of programming experience in C. A desire to write a feature for major software like Scribus from scratch is my 'Developer's Dream'. I want to fulfill it and also help to improve Scribus. Interacting with the Scribus team has been an absolute pleasure. Plus, the 'Scribus Culture' has impressed me so much that I am no longer trying to apply to any other GSoC organizations.

Overview

After discussing my plans with the Scribus developers, the following goals are to be met by the time of completion of the project:

Rewrite the source code to make it more modular starting with separation of GUI from the search logic.
Implement support for regular expressions in the search using Qregex.
Implement simultaneous search and replace in multiple documents
Improve user interface by separating Basic Search and Advanced Search.

Implementation

Essential

Following features have to be implemented by end of GSoC 2009:-

The new 'Search and Replace' will have two 'Search and Replace' dialogs. The first one would search in a particular text region. The second will offer the user to search the complete document or selected parts of it. It will first categorize the entire Scribus document to enumerate its individual parts and then invoke the first 'Search and Replace' on all or only selected parts.
The new 'Search and Replace' dialog will have two modes. One will perform basic 'Search and Replace' using text based matching. The second mode will be the full fledged dialog similar to what we have today. The second one can be invoked from the first by clicking on the 'Advanced Search' button or can be invoked directly.
New Search and Replace will support Regular Expressions. That means we can for instance find all words beginning with or containing a particular string or character at a particular location.
The code will be divided into two components - the GUI and the actual code logic. This would allow 'Search and Replace' to be extended much more easily later on. As a feature might require an addition to the UI part and a separate code backend, changes in one can be done independently of the other.

Optional

If any time is left, I would also like to implement the following ideas:

To make the 'Search and Replace' dialog more informative, proper message boxes would be provided such as 'Match not Found' unlike the current austere 'Search Complete' message.
Depending on the 'textbox' support by Qt, I could add an incremental search capability. This feature would need a mechanism in text frames to send the characters as soon as they are being typed rather than only after the 'Return' key is pressed.
To have 'Search and Replace' automatically take input from the text highlighted by user when it is invoked (From the bugtracker).
To have a 'Warp Around' feature. If a user enables it then when the end of the document is reached, search starts from the beginning again. This would be needed only for the search component.
To shift the 'Search and replace' dialog aside when a user needs to view the documents being searched.

Scribus Users Feedback

Please dont forget the 'style' searches :

searching occurences of 'needle' with bold characters
searching occurences of text with font soandso
searching occurences of 'needle' with font size = 10
searching occurences of 'needle' with Title1 style
searching occurences of 'needle' with not-bold characters
searching occurences of 'needle' with characters (either bold or not-bold) AND italic
etc... with all possible mix, including negative criteria and unknown values (3 states buttons style-search-specs)

idem with replacing :
replace style with bold and not italic
replace style with bold but dont modify italic or not-italic (3 states buttons style-replace-spects)

Timeline

May 23 – June 10: Basic 'Search and Replace' without Regular Expressions
June 10 – June 23: Basic 'Search and Replace' with Regular Expressions
June 23 - July 6: Advanced 'Search and Replace'
July 6 – July 21: Global 'Search and Replace'
July 21 - August 10: Exploiting bugs and patching them; implementation of additional ideas if time permits.
August 10 - August 17: Documenting the source code; writing about other ideas that can be implemented.

Deliverables

A new "Search and Replace" UI with all the features mentioned under the "Essential" portion of “Implementation”.
Project documentation.
Some or all features from the “Optional" portion of “Implementation”

BIO

Meetanshu Gupta : meetanshu.gupta@gmail.com

I am a final year Undergraduate student in Computer Engineering at the Malaviya National Institute of Technology, Jaipur, India. I have been programming in C since Class 6 and C++ since Class 11. I have worked on Qt2 to create a UI for the C++ project "Department Management Software." I have done many projects in C, C++ for academics, competitions and at college level. My interests include programming (C/C++/Java/C#), algorithms, network programming. Last summer I had a two-months internship at Microsoft IDC, Hyderabad, India and got some hands-on experience on real world projects. Since then I have taken interest in reading and understanding Open Source codebases (Scribus, Vim, OpenOffice.org), learning debugging techniques and amicably working in teams.