Thank you for the insight and the references! Quite useful to learn about this stuff.
вс, 5 апр. 2020 г. в 07:46, Dmitry Gutov <dgu...@yandex.ru>: > > Hi! > > Some late clarifications about mmm-mode. > > On 26.04.2019 00:00, Dmitrii Korobeinikov wrote: > > > A lot of the functionality of MMM Mode---that which makes the major mode > > > appear to change---is implemented by saving and restoring the values of > > > local variables, or pseudo-variables. > > > > What I don't understand is where the modes of the submode region run and > > when they are turned on. > > They are run in an empty temporary buffer, see mmm-update-mode-info. > That is true for all the "submodes" in a buffer. The primary major mode > is run in the context of that buffer (IIRC). After any of them runs, the > code responsible for it collects the values of a certain number of known > variables and associates that map with the major mode (this is a bit of > a simplification). > > > Are necessary modes just allowed to run at the right time for the whole > > buffer? > > When you move between the "chunks", no major mode functions are called. > Instead, the values of variables are swapped in. Including the value of > the 'major-mode' variable. > > > But then, how are they limited in their effect to just the > > necessary region? Narrowing? > > Usually, yes. Especially when we're talking about font-lock and > syntax-propertize-function. See mmm-fontify-region-list for an example. > > > Could, for example, syntax checking be done efficiently that way? > > That depends on the combination of modes and how they are used (either > they can be nested, like in web templates, or it's a flat list where > chunks are largely independent like in Jupyter). But in most cases, I > think, you could pick a good strategy. There are no universal ones, though.
