Difference between revisions of "OSGeo Cartographic Engine"

From OSGeo
Jump to navigation Jump to search
 
(99 intermediate revisions by 30 users not shown)
Line 1: Line 1:
== Motivation ==
+
'''Summer of Code 2010:  This project has been proposed for [http://wiki.osgeo.org/wiki/Google_Summer_of_Code_2010_Ideas#The_ideas_pages SoC 2010]'''
  
There are plans about new cartographic tools (map composer) for QGIS, as well as for GRASS. We suggest to develop a joint library for the various functions which could be shared across GRASS, QGIS and other projects with bindings to Qt, Python etc. The GUI part will be done within the individual software projects.
+
'''JOIN MAILING LIST''': A list is now available for discussion around these concepts/ideas: [http://lists.osgeo.org/mailman/listinfo/carto Carto list].
 +
-----
 +
'''NOTE: This is not actually about building an API, rather about establishing some specifications for interaction between some common applications and a map production engine. '''
 +
-----
 +
[[Image:Carto_engine_v1.png | thumb | right | 300px | Conceptual overview]]
  
== Functionality ==
+
== Motivation ==
 
+
To develop a common method and tools for printing high quality maps from a variety of mapping software.  Starting with Mapnik as the renderer and QGIS, GRASS or others as the source application. Initial focus on building standards, then building command line tools or GUI tools.
Shared functions may include
 
 
 
* class breaks (various algorithms)
 
* histogram analysis
 
* color ramps
 
* RGB and possibly CYMK support (aka color transformations) etc.
 
* Algorithms for intelligent label placement
 
* Symbology management
 
* map surround
 
* legend
 
* cartographic templates (ie. predefined map styles)
 
 
 
== Programming Language ==
 
  
* C? C++?
+
There are plans for new/rewritten cartographic tools (map composer) for QGIS, as well as for GRASS. We suggest to develop a joint library for the various functions which could be shared across GRASS, QGIS and other projects with bindings to Qt, Python etc.  The concept applies to all OSGeo projects in general as well - so high quality production can be something all projects have access to without having to code it from scratch in their own applications.
  
Perhaps we could implement the core in C++, but provide a high level C API for applications and environments wanting to drive the libraryC++ is a strong implementation language, but presenting a C API to the outside world reduces complexity calling from other languages or compiler versions and reduces "interface fragility".  
+
This applies to web mapping applications as well as desktop ones where high resolution, large format printing is requiredBut it also applies generally to smaller print jobs as well, e.g. high resolution laser printers.
  
== Output Formats ==
+
The GUI part for exporting layouts, etc. could be done within the individual software projects, while this project focuses on standard libraries or command line utilities to ''process'' compliant documents.  The production of a printing interchange specification/standard is needed as well.  An advanced layout tool in the form of a GUI would also be very useful. 
  
What output formats do we want to targetI think at the core this question comes down to whether we want to render everything to a raster format or whether we wish to preserve linework, and text in non-raster form.  That is, whether output output is a raster format, or a raster+vector+text format like postscript/pdf.
+
The concepts and idea below are quite diverse and represent a broad range of inputIf a SoC student takes on this project, part of their job would be to help focus on a particular part of the problem (and possible clarify some of the others into workable sized components).
  
If a cartographic product is the objective, I think selection of an output format like PDF makes a lot of sense, but it does make interactive composition more complicated and it is possible we have limited pdf/postscript development expertise in our developer community.
+
== Ideas for Summer of Code 2010 Project ==
  
== Existing code to be recycled ==
+
This year the project will focus on building the capabilities using the [http://mapnik.org Mapnik] package due to its ability to output high quality vector maps.  It supports XML based configuration files, but needs additions to handle the needs of this project.  The QGIS [http://bitbucket.org/springmeyer/quantumnik/wiki/Home Quantumnik plugin] is an excellent example of moving toward the end goal of this project.  Other applications may be considered too.
  
There will be code in GRASS, QGIS, the Python Cartographic Library (http://zcologia.org/cartography) and other sources which could be integrated (check license compatibility). Please indicate precisely where to find the source code, license, programming language:
+
Some further ideas to consider (add your own!!):
 +
* An approach for transforming XML files from other applications into format compatible with Mapnik or this project specs
 +
* How to handle caching web services at the proper resolution for printing?
 +
* Consider icon collision handling
 +
* Tweaking Mapnik and underlying Cairo library to handle text/font exporting better
 +
* ...what else....?
  
* Label placement code:
+
A workplan might look like this, but is totally open to new ideas:
** Mapserver [http://trac.osgeo.org/mapserver/browser/trunk/mapserver/maplabel.c labeling code] (MapServer License, C language)
+
** GRASS [http://trac.osgeo.org/grass/browser/grass/trunk/vector/v.label.sa v.label.sa] (GPL, C language)
+
# Review existing map definition standards (e.g. from a few desktop and web applications [[Configuration Files]])
* Color management:
+
# Identify missing components from those standards.
** MapGuide, [http://trac.osgeo.org/mapguide/browser/trunk/MgDev/Server/src/Services/Feature/FeatureNumericFunctions.cpp CalculateDistribution] and [http://trac.osgeo.org/mapguide/browser/trunk/MgDev/Common/PlatformBase/Data/Color.cpp Color] (LGPL.C++)
+
# Create new specification or make a choice of existing one. Document the primary components as necessary for other app developers to build to its specification.
** OpenJump ?
+
# Identify print-specific options/configuration (and challenges) needed for producing a hard copy. (see Potential Components below for an idea of overall concept)
* Stylization:
+
# Write scripts to consume a compliant document and produce a output file - e.g. postscript, PDF or raster. (See discussion below re: output formats)
* Postscript/PDF output:
+
# If possible, write a sample GUI for one OSGeo project to use the specification.
** GRASS [http://trac.osgeo.org/grass/browser/grass/trunk/ps/ps.map ps.map] (Postscript output) and [http://trac.osgeo.org/grass/browser/grass/trunk/lib/psdriver PS driver]
 
* PNG output:
 
** GRASS [http://trac.osgeo.org/grass/browser/grass/trunk/lib/pngdriver PNG driver]
 
** See AGG below
 
  
* Class breaks:
+
== Potential Components ==
** GRASS [http://trac.osgeo.org/grass/browser/grass/trunk/lib/arraystats different algorithms for class breaks] (GPL, C)
+
Breaking the end goal into some components will help to recognise which area(s) there is greatest need. There seems to be at least four primary components to the proposal. Any of these would be worth looking at, but it is not necessarily a final list.  Add your ideas!
  
== Technologies ==
+
'''Style & Layout Configuration Standards'''
 +
* Specifications for styling of features and layout of a map composition
 +
* Could re-use existing approaches where applicable (e.g. SLD, WMC, MapServer Mapfile, OGR Feature Styles)
 +
* Specifications to be used in map definition/configuration files
  
* AGG (sophisticated rendering of vector/text graphics to raster format - used by MapServer, MapGuide, Mapnik).
+
'''Graphical Style & Layout Editor'''
* Freetype - text management and rendering library
+
* To create styles interactively
 +
* To define map layout
 +
* Could be stand-alone
 +
* Could be adopted as part of another application
 +
* Saves a style/layout definition document according to above standards
  
== License ==
+
'''Graphical Map Composition Editor'''
 +
* Implemented by existing applications (e.g. GRASS, QGIS, etc.)
 +
* Sets up pointers to data layers
 +
* Assigns styles definitions to the layers
 +
* Applies layout definition to maps
 +
* Outputs a map composition configuration file
  
It is proposed to license the OSGeo Cartographic Library under LGPL. The license needs to be GPL compliant.
+
'''Rendering Engine'''
 +
* Stand-alone application (no GUI)
 +
* Reads standard map composition configuration file
 +
* Reads and applies the required style/layout definition documents
 +
* Outputs final graphics to specific files
  
== Interested people ==
+
== See also ==
  
* [[User:Neteler|Markus Neteler]] (proposer)
+
* [[OSGeo Cartographic Engine Discussion]] - more (original) detailed brainstorming and ideas
* Horst Düster
+
* [[OSGeo map symbol set]]
* [[User:Landa|Martin Landa]]
+
* [[Configuration Files]]
* [[User:MarcoHugentobler|Marco Hugentobler]]
+
* PAL - A cartographic labelling library ([http://pal.heig-vd.ch website], FOSS4G2008 [http://foss4g.org/index.php/foss4g/2008/paper/view/275 abstract], [http://mistic.heig-vd.ch/taillard/articles.dir/pal_foss2008.pdf PDF])
* Frank Warmerdam
+
* [[Programming Languages Project]]
* [[User:Adiez|Agustin Diez]]
 
* [[User:mlennert|Moritz Lennert]]
 
* [[User:Dsampson | Dave Sampson]]: End User, Search and Rescue Maps,
 
  
 
[[Category: Software Stack]]
 
[[Category: Software Stack]]
 +
[[Category: Printing]]

Latest revision as of 23:42, 6 April 2010

Summer of Code 2010: This project has been proposed for SoC 2010

JOIN MAILING LIST: A list is now available for discussion around these concepts/ideas: Carto list.


NOTE: This is not actually about building an API, rather about establishing some specifications for interaction between some common applications and a map production engine.


Conceptual overview

Motivation

To develop a common method and tools for printing high quality maps from a variety of mapping software. Starting with Mapnik as the renderer and QGIS, GRASS or others as the source application. Initial focus on building standards, then building command line tools or GUI tools.

There are plans for new/rewritten cartographic tools (map composer) for QGIS, as well as for GRASS. We suggest to develop a joint library for the various functions which could be shared across GRASS, QGIS and other projects with bindings to Qt, Python etc. The concept applies to all OSGeo projects in general as well - so high quality production can be something all projects have access to without having to code it from scratch in their own applications.

This applies to web mapping applications as well as desktop ones where high resolution, large format printing is required. But it also applies generally to smaller print jobs as well, e.g. high resolution laser printers.

The GUI part for exporting layouts, etc. could be done within the individual software projects, while this project focuses on standard libraries or command line utilities to process compliant documents. The production of a printing interchange specification/standard is needed as well. An advanced layout tool in the form of a GUI would also be very useful.

The concepts and idea below are quite diverse and represent a broad range of input. If a SoC student takes on this project, part of their job would be to help focus on a particular part of the problem (and possible clarify some of the others into workable sized components).

Ideas for Summer of Code 2010 Project

This year the project will focus on building the capabilities using the Mapnik package due to its ability to output high quality vector maps. It supports XML based configuration files, but needs additions to handle the needs of this project. The QGIS Quantumnik plugin is an excellent example of moving toward the end goal of this project. Other applications may be considered too.

Some further ideas to consider (add your own!!):

  • An approach for transforming XML files from other applications into format compatible with Mapnik or this project specs
  • How to handle caching web services at the proper resolution for printing?
  • Consider icon collision handling
  • Tweaking Mapnik and underlying Cairo library to handle text/font exporting better
  • ...what else....?

A workplan might look like this, but is totally open to new ideas:

  1. Review existing map definition standards (e.g. from a few desktop and web applications Configuration Files)
  2. Identify missing components from those standards.
  3. Create new specification or make a choice of existing one. Document the primary components as necessary for other app developers to build to its specification.
  4. Identify print-specific options/configuration (and challenges) needed for producing a hard copy. (see Potential Components below for an idea of overall concept)
  5. Write scripts to consume a compliant document and produce a output file - e.g. postscript, PDF or raster. (See discussion below re: output formats)
  6. If possible, write a sample GUI for one OSGeo project to use the specification.

Potential Components

Breaking the end goal into some components will help to recognise which area(s) there is greatest need. There seems to be at least four primary components to the proposal. Any of these would be worth looking at, but it is not necessarily a final list. Add your ideas!

Style & Layout Configuration Standards

  • Specifications for styling of features and layout of a map composition
  • Could re-use existing approaches where applicable (e.g. SLD, WMC, MapServer Mapfile, OGR Feature Styles)
  • Specifications to be used in map definition/configuration files

Graphical Style & Layout Editor

  • To create styles interactively
  • To define map layout
  • Could be stand-alone
  • Could be adopted as part of another application
  • Saves a style/layout definition document according to above standards

Graphical Map Composition Editor

  • Implemented by existing applications (e.g. GRASS, QGIS, etc.)
  • Sets up pointers to data layers
  • Assigns styles definitions to the layers
  • Applies layout definition to maps
  • Outputs a map composition configuration file

Rendering Engine

  • Stand-alone application (no GUI)
  • Reads standard map composition configuration file
  • Reads and applies the required style/layout definition documents
  • Outputs final graphics to specific files

See also