Abdelrazak Younes <[EMAIL PROTECTED]> writes:
| Abdelrazak Younes wrote:
| > Abdelrazak Younes wrote:
| >>
| >> This patch does that and it removes absolutely _no_ functionality.
| >> Actually, I think it should speed-up _significantly_ the latex
| >> compilation for example in case of problematic files.
| > I am going tomorrow and I really don't have the time to wait for
| > Lars approval. I am going to commit this because I think this is the
| > right thing to do. If you are not happy with it, just revert the
| > patch(es), I won't be offended ;-)
| > Lars, before you get upset about the transfer of errorList_ to
| > Buffer please have a little faith in me. I agree this is not the
| > right place to put it but this is transitional. Transferring to
| > Buffer simplify the code and the work flow quite significantly with
| > minimal changes to the code. My next patch(es) will modify that.
| > I will then proceed to step 1 or maybe to steps 1 and 2 at the same
| > time. I will probably use a global ErrorLists map variable. I hate
| > global variables but considering the current structure of the code I
| > have no real choice for now. I'll try to find some time to implement
| > the one (theApp) unique variable when I come back.
| > Let's see how fast I can code the patch.
|
| Hum... 2 hours... I need some sleep!
|
| Attached my current patch. It works well at least for the loading of
| buggy documents (like document with no \begin_document). I haven't
| tested yet LateX compilation. There might be some adjustments to do
| WRT ErrorLists initialisation but the big work is done.
|
| The good news is that, with this infrastructure, the error lists are
| persistent. With a few more code, it should be possible to show the
| LateX processing errors even after the error dialog has been closed.
I think this is the right approach, but (always a but ;-)) I don't
think we should use a string as index. Also the globalness of the list
structure is not good. All buffers, conversion errrors will get mashed
into the same structrue, so a need to lookup per buffer is there.
--
Lgb
