CSL processor news

CSL Development
21

…

It’s just a sample, you can write tests for larger blocks of CSL with
the same layout. Fine-grained tests like this of individual elements
and attributes are helpful to a coder, though, to provide a target for
specific functionality, and to identify specific areas of misbehaviour
when later changes to a program mess things up.

Right, but the results examples are a) specific to one output format
(HTML) and b) could be represented many ways in that format and still
be correct. E.g. rather than …

“some title”

… you could have:

“some title”

… or:

“some title”

… or:

“some title”

All of them would yield the same display.

So am simply observing it’d be nice to be able to operate a level
above that if possible to avoid that problem.

Bruce

22

Do you have any idea what this level would be? It agree that this is a
real issue (especially if you were to have something like ), but I have no
inspiration as far as good ways of handling it. The three cases you’ve
mentioned could all be accommodated by easily with simple string
replace operations, but the above case would be a little more difficult.

Simon

23

…

It’s just a sample, you can write tests for larger blocks of CSL with

the same layout. Fine-grained tests like this of individual elements

and attributes are helpful to a coder, though, to provide a target for

specific functionality, and to identify specific areas of misbehaviour

when later changes to a program mess things up.

Right, but the results examples are a) specific to one output format
(HTML) and b) could be represented many ways in that format and still
be correct. E.g. rather than …

We’re just trying to confirm that the engine is capable of doing the
right thing. The markup can be generated by a driver specific to the
testing process, and it can contain any symbols you like. I used HTML
just because it’s familiar to me, but if something else should be
used, no problem.

“some title”

… you could have:

“some title”

… or:

“some title”

… or:

“some title”

All of them would yield the same display.

So am simply observing it’d be nice to be able to operate a level
above that if possible to avoid that problem.

Do you have any idea what this level would be? It agree that this is a real
issue (especially if you were to have something like ), but I have no inspiration as
far as good ways of handling it. The three cases you’ve mentioned could all
be accommodated by easily with simple string replace operations, but the
above case would be a little more difficult.

Can you combine tags like that in RTF? If not, I think you can safely
assume the harder case away (we might check the existing
implementations just to be sure no one is getting burned). What would
be a problem is a markup language that requires formatting hints to be
combined in this way, and does not permit them to be nested. I can’t
think of any, but my knowledge is limited.

Frank

24

…

It’s just a sample, you can write tests for larger blocks of CSL with

the same layout. Fine-grained tests like this of individual elements

and attributes are helpful to a coder, though, to provide a target for

specific functionality, and to identify specific areas of misbehaviour

when later changes to a program mess things up.

Right, but the results examples are a) specific to one output format
(HTML) and b) could be represented many ways in that format and still
be correct. E.g. rather than …

We’re just trying to confirm that the engine is capable of doing the
right thing. The markup can be generated by a driver specific to the
testing process, and it can contain any symbols you like. I used HTML
just because it’s familiar to me, but if something else should be
used, no problem.

“some title”

… you could have:

“some title”

… or:

“some title”

… or:

“some title”

All of them would yield the same display.

So am simply observing it’d be nice to be able to operate a level
above that if possible to avoid that problem.

Do you have any idea what this level would be? It agree that this is a real
issue (especially if you were to have something like ), but I have no inspiration as
far as good ways of handling it. The three cases you’ve mentioned could all
be accommodated by easily with simple string replace operations, but the
above case would be a little more difficult.

Can you combine tags like that in RTF? If not, I think you can safely
assume the harder case away (we might check the existing
implementations just to be sure no one is getting burned). What would
be a problem is a markup language that requires formatting hints to be
combined in this way, and does not permit them to be nested. I can’t
think of any, but my knowledge is limited.

There might be some engines that don’t control the order of nesting
for some decorations:

My Title
My Title

For this, and for Simon’s harder case, I guess alternative valid
responses could be specified in the JSON by enclosing a set of results
in an array.

25

…

There might be some engines that don’t control the order of nesting
for some decorations:

My Title
My Title

For this, and for Simon’s harder case, I guess alternative valid
responses could be specified in the JSON by enclosing a set of results
in an array.

Yeah.

I guess we just need to ask, what do we need to test for output? For a
test on styling, can we not just test for only that?

Similarly, we can have tests that only check other specific output things:

  • field order
  • substitution
  • year-suffix generation and other disambiguation
  • quote handing
  • name display
  • punctuation and whitespace

Bruce

26

…

There might be some engines that don’t control the order of nesting
for some decorations:

My Title
My Title

For this, and for Simon’s harder case, I guess alternative valid
responses could be specified in the JSON by enclosing a set of results
in an array.

Yeah.

I guess we just need to ask, what do we need to test for output? For a
test on styling, can we not just test for only that?

Similarly, we can have tests that only check other specific output things:

  • field order
  • substitution
  • year-suffix generation and other disambiguation
  • quote handing
  • name display
  • punctuation and whitespace
27

Hi Frank,
Whose “summer”; the one down south, or up north? So June, or December?

He teaches in Japan, so the one north… June, I believe.

Frank, thanks for your contribution, I’m looking forward to seeing you
back here hacking again. Its’ been a pleasure getting to know I’m not
the only comparative lawyer wasting her time and jeopardizing an
academic career just for the joy of coding. :wink:

Andrea,

The feeling is mutual, I’ve been very glad to discover your work.
I’ll be following the list in the next months, I’m just taking a
holiday from work on the program. This morning I did some final
tidying up in the code and comments, and the online version gives a
very clear picture of how things fit together.

The programmers that built the English website of our faculty
(http://gsl-nagoya-u.net) are big on Haskell and Ruby. Their main
line of work is not connected with biblio management in any way, but
their site is here if you’re interested, with a blog (My Beautiful
Code) where they occasionally post on software process and design:
http://www.starling-software.com/en/

Frank