Fix for issue #30

[David-Polehonski]

Added editor listeners to set .cfm files to html/cfml mixed mode if the file contains non cfml tags.

[dajester2013]

I would think it would be better to have the html grammar added back to cfml.cson, as cfml tags are intended to be interspersed with html.

The only thing I don't like about having to scan for mixed mode is if you run into a situation where you are scanning hundreds of lines on every change to the file (the event is slightly buffered, but every keystroke runs the did-change event).

[David-Polehonski]

@dajester2013 We could only scan the file on open? I wouldn't remove the cfml only mode, but we could remove the fileTypes array? the cfml only mode allows for a separation of concerns and I find it particularly useful when testing.

[dajester2013]

Scanning the file on change allows someone to open a blank cfc file with the CFML grammar active, then start typing component and have it switch automatically to CFScript. You could assume that the default for cfc's is cfscript, except when someone starts typing <cfcomponent>, it would not automatically switch over to cfml if the scanning was only on file open.

I was reserved on adding the file change listener to begin with, that's why I only looked for the first non-empty line.

[David-Polehonski]

I just peaked at the docs for the did-change, it passes a delta containing the line/s that have been changed. Think we could write it to only update based on the actual changed lines? I.e if they type <cfcomponent on line four the delta reads {'start':4, 'end':4} so we could just read that line to trigger a switch?

[dajester2013]

I don't know if that would work. The grammar should only be determined based on the first line.

I originally used a regex scan, but I had a problem where if I typed something like:

import a.b.c;

<cfcomponent>

it would start switching back and forth between the cfml and cfscript grammars rapidly. I think that would be the case if it were only inspecting the deltas.

[David-Polehonski]

Well removing the fileTypes array seems to have done the trick regarding the HTML mixed mode - at least it seems to be working for me. With that being the case then I think your initial first line scan should be sufficient to most use cases (Read: I can't think of one where it fails).

[atuttle]

Sorry, I'm quite busy today and having some trouble keeping up with all of the conversations. Is this PR ok to merge, then?

[dajester2013]

I'm good with it.
On Jul 7, 2015 1:05 PM, "Adam Tuttle" notifications@github.com wrote:

Sorry, I'm quite busy today and having some trouble keeping up with all of
the conversations. Is this PR ok to merge, then?

—
Reply to this email directly or view it on GitHub
#35 (comment)
.

[atuttle]

$ apm publish patch
Preparing and tagging a new version ✓
Pushing v0.8.3 tag ✓
Publishing language-cfml@v0.8.3 ✓