General suggestion: Unit tests for ConText
Hi, during my work with context I wrote a bunch of fewliners mainly to test how to do things. When there change from mkii to mkiv occured many of my mini-tests failed and later with the growing of mkiv restartet to work - some just without adaption.. some with some adjustment to a new syntax. I realize that they have the same functionality like unit-tests - just instead of comparing classes and variables - a pdf-output is visually compared. So I like to suggest to introduce unit tests for context. It would a) increase the quality of context as it makes the testing easier for the developer b) it also would be a great source of minihowtos for learning context. Before I write my technical ideas about this I would like to know what are you thinking about this idea?
Dear Peter, Am Montag, den 18.07.2011, 19:51 +0200 schrieb Peter Schorsch: […]
I realize that they have the same functionality like unit-tests - just instead of comparing classes and variables - a pdf-output is visually compared.
So I like to suggest to introduce unit tests for context. It would
a) increase the quality of context as it makes the testing easier for the developer
b) it also would be a great source of minihowtos for learning context.
Before I write my technical ideas about this I would like to know what are you thinking about this idea?
not being a developer I like that idea very much. Could the PDF output be converted to a picture and some routine be used to determine a difference between the template picture and the created one? Thanks, Paul
On 18-7-2011 7:51, Peter Schorsch wrote:
Hi,
during my work with context I wrote a bunch of fewliners mainly to test how to do things. When there change from mkii to mkiv occured many of my mini-tests failed and later with the growing of mkiv restartet to work - some just without adaption.. some with some adjustment to a new syntax.
I realize that they have the same functionality like unit-tests - just instead of comparing classes and variables - a pdf-output is visually compared.
So I like to suggest to introduce unit tests for context. It would
a) increase the quality of context as it makes the testing easier for the developer
b) it also would be a great source of minihowtos for learning context.
Before I write my technical ideas about this I would like to know what are you thinking about this idea?
fyi: there is an archive with a bunch of tests (mostly mine but I occasionally add -adapted- samples from the list): http://www.pragma-ade.nl/context/latest/cont-tst.7z Quite some of these can be used for testing, some can also serve as examples of usage. I have a script that can process them and keep track of changes at a rough pixel level but simply have no time to keep up with that. Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------
2011/7/19 Hans Hagen
fyi: there is an archive with a bunch of tests (mostly mine but I occasionally add -adapted- samples from the list):
http://www.pragma-ade.nl/**context/latest/cont-tst.7zhttp://www.pragma-ade.nl/context/latest/cont-tst.7z
Quite some of these can be used for testing, some can also serve as examples of usage. I have a script that can process them and keep track of changes at a rough pixel level but simply have no time to keep up with that.
How much time would cost that? Maybe I could do that. -- Cecil Westerhof
2011/7/18 Peter Schorsch
during my work with context I wrote a bunch of fewliners mainly to test how to do things. When there change from mkii to mkiv occured many of my mini-tests failed and later with the growing of mkiv restartet to work - some just without adaption.. some with some adjustment to a new syntax.
I realize that they have the same functionality like unit-tests - just instead of comparing classes and variables - a pdf-output is visually compared.
So I like to suggest to introduce unit tests for context. It would
a) increase the quality of context as it makes the testing easier for the developer
b) it also would be a great source of minihowtos for learning context.
Before I write my technical ideas about this I would like to know what are you thinking about this idea?
I think it is a great idea. -- Cecil Westerhof
participants (4)
-
Cecil Westerhof -
Hans Hagen -
Paul Menzel -
Peter Schorsch