Newly created files in Atom are always "Plain Text". How can I change this so that new files will automatically be in another language, for example "Shell Script (Bash)"? I want to do this because auto indentation does not work with Plain Text files.
Had this problem as well, there is a plugin called default-language that will do this for you.
Search atom for default-language, install and open its settings. Type the name of the language you want Atom to default to, e.g. Shell Script (if in doubt, copy from the language selection menu) in the Default Language field. Next time you open a script with no extension (or shebang) it'll default to the language you set.
The following code, added to your init.coffee, will do what you're asking:
atom.workspace.observeTextEditors (editor) ->
default_scope = 'source.shell'
original = editor.getGrammar()
# If the editor has "null" grammar (aka unset)
if original? and original is atom.grammars.grammarForScopeName('text.plain.null-grammar')
default_grammar = atom.grammars.grammarForScopeName(default_scope)
if default_grammar? # check if desired grammar is already loaded
editor.setGrammar(default_grammar)
else
# grammar was not loaded yet, so add a callback as grammars load
callback = atom.grammars.onDidAddGrammar (grammar) ->
if grammar.id is default_scope
# once we've loaded the grammar, set it and dispose of the callback
editor.setGrammar(grammar)
callback.dispose()
Things to note:
The init.coffee file is where you can customize Atom without having to write a package
The observeTextEditors method sets a callback that is called upon each TextEditor creation for currently open and future editors
The code above:
Checks the grammar that the editor was created with
If and only if it is the default ("null") grammar, it sets the editor's grammar to the Shell grammar once it's loaded
Disposes of the callback to check for grammar loading once it's done with it
This should solve the TypeError: Cannot call method 'getScore' of undefined that happens for the first file opened in a new window.
To default to a different grammar, just change the default_scope = 'source.shell' line to use the scope of whatever grammar you'd like.
Firstly, CTRL+SHIFT+L is your friend. It's unfortunately not a permanent solution, but nice to know about.
Of course, we'd prefer a more permanent solution. A couple of the other answers are now obsolete due to API changes in Atom. Below is a more up-to-date version. Inspiration initially came from this discussion, but the other answers here seem to follow the same concept as well.
Place this in your init.coffee file (File -> Open Your Init Script):
extname = require("path").extname
fileTypes =
".wxs": "text.xml"
".wxi": "text.xml"
".wixobj": "text.xml"
nullGrammar = atom.grammars.selectGrammar("text.plain.null-grammar")
atom.workspace.observeTextEditors (editor) ->
grammar = atom.grammars.selectGrammar(fileTypes[extname(editor.getPath())])
editor.setGrammar grammar if editor.getGrammar() is nullGrammar and grammar isnt nullGrammar
Basically, you define have an array of file types, and the grammars (AKA syntax highlighting) that you want to associate them with. Find each editor, find out if it has a selected grammar already, and if not, attempt to give it one if we find one.
The one issue I've had with this is that the syntax highlighting only works if you open files after you've already launched Atom; if you open a file that results in Atom launching (say by double clicking on it's icon in your favourite OS), syntax highlighting won't for that file until you re-open it.
You need to create a mapping in your config.cson file.
"*":
core:
customFileTypes:
"source.shell": [
"sh"
"shell"
]
For mapping .sh and .shell files to shell script syntax.
Have a look at this bit of code: (you can can then change 'text.html.basic' to whichever syntax you require)
editor = atom.workspace.getActiveTextEditor()
cursor = editor.getLastCursor()
valueAtCursor = atom.config.get(cursor.getScopeDescriptor(), 'my-package.my-setting')
valueForLanguage = atom.config.get(editor.getRootScopeDescriptor('text.html.basic'), 'my-package.my-setting')
For reference please see: Scope Descriptors # https://atom.io/docs/latest/advanced/scopes-and-scope-descriptors
Related
I'm running a u-sql script similar to this:
#output =
SELECT Tag,
Json
FROM #table;
OUTPUT #output
TO #"/Dir/{Tag}/filename.txt"
USING Outputters.Text(quoting : false);
The problem is that #output is empty and the execution crashes. I already checked that if I don't use {tag} in the output path the script works well (it writes an empty file but that's the expectable).
Is there a way to avoid the crash and simply don't output anything?
Thank you
The form you are using is not yet publicly supported. Output to Files (U-SQL) documents the only supported version right now.
That said, depending on the runtime that you are using, and the flags that you have set in the script, you might be running the private preview feature of outputting to a set of files. In that case, I would expect that it would work properly.
Are you able to share a job link?
In my scripts, which I am currently making into a package, a number of "global settings" are needed. Currently, these settings are in global variables and were usually changed by editing the script directly. (The script produces entries for a database, and you need to adjust stuff like "author name" and other custom "constant" part of the entries.)
Again, currently I used const_author <- "Meow The Scientist Cat"et al. I can, of course, leave this exactly as is, and export all the global variables, so the user can set them to whatever. However, this is ugly, and pollutes the namespace.
What is the standard method in R to make such settings available to the user? Using options()? And at which point in the package should these options be loaded?
Maybe using a function like settingsTemplate(filename) which exports a file with default settings, which the user then can customize; and he has to source the file or loadSettings (filename) before using the scripts?
You could create something similar to xcms: in zzz.R we call .setXCMSOptions (from init.R upon package loading, where xcms specific options are inserted into the generic BioC options:
getOption("BioC")$xcms
You could provide getter and setter methods for your options.
I've been trying to figure out how to programmatically add files to an Xcode4 project and it seemed like AppleScript would be the way to go, however I'm running into "missing value" errors.
Here's the code I've got:
tell application "Xcode"
set theProject to first project
set theTarget to first target of theProject
set theBuildPhase to compile sources phase of theTarget
tell first group of theProject
set theFileRef to make new file reference with properties {full path:"/Users/jeff/Projects/XcodeTest/XcodeTest/MyViewController.h", name:"MyViewController.h", path:"XcodeTest/MyViewController.h", path type:group relative}
add theFileRef to theProject
end tell
--tell theBuildPhase to make new build file with properties {build phase:theBuildPhase, name:"MyViewController.h", file reference:theFileRef, target:theTarget, project:theProject}
end tell
I've tried the commented-out line instead of the add-command as well, but that doesn't work either (I get "missing value").
The 'add' error is:
error "Xcode got an error: file reference id \"251AD3431432472E006E300F\" of Xcode 3 group id \"251AD32C14324645006E300F\" of project \"XcodeTest\" of workspace document \"XcodeTest.xcodeproj/project.xcworkspace\" doesn’t understand the add message." number -1708 from file reference id "251AD3431432472E006E300F" of Xcode 3 group id "251AD32C14324645006E300F" of project "XcodeTest" of workspace document "XcodeTest.xcodeproj/project.xcworkspace"
The "make new reference" does add the file to the list of files in Xcode, but I also need it to be added to the project target so that I can add actions and outlets to the file from Xcode w/o having to first check the checkbox to add it to the "target membership".
I ended up sending this question to the devs on the xcode developer list and the response I got was effectively "you can't".
This appears to be completely broken in Xcode4, but I've seen a project that does it. I think what they are doing is parsing and modifying the "project.pbxproj" file directly. (this file is hidden inside the xcodeproj bundle)
The file is a GUID soup, but once you look at it for a while it seems possible to safely modify it, especially if you are only adding stuff.
Edit:
Found this stackoverflow answer that might help.
Tutorial or Guide for Scripting XCode Build Phases
There is a poorly documented user defined build setting that can be added. Files can be both excluded and included from compilation
Go to your target's Build Settings > Tap the + button > Add User-Defined Setting
The key is either INCLUDED_SOURCE_FILE_NAMES or EXCLUDED_SOURCE_FILE_NAMES
The value is a space separated list of file paths
See reference:
http://lists.apple.com/archives/xcode-users/2009/Jun/msg00153.html
A bit of a followup from a previous question.
As I mentioned in that question, my overall goal is to call a Ruby script after ImageCache does its magic with generating thumbnails and whatnot.
Sebi's suggestion from this question involved using hook_nodeapi.
Sadly, my Drupal knowledge of creating modules and/or hacking into existing modules is pretty limited.
So, for this question:
Should I create my own module or attempt to modify the ImageCache module?
How do I go about getting the generated thumbnail path (from ImageCache) to pass into my Ruby script?
edit
I found this question searching through SO...
Is it possible to do something similar in the _imagecache_cache function that would do what I want?
ie
function _imagecache_cache($presetname, $path) {
...
...
// check if deriv exists... (file was created between apaches request handler and reaching this code)
// otherwise try to create the derivative.
if (file_exists($dst) || imagecache_build_derivative($preset['actions'], $src, $dst)) {
imagecache_transfer($dst);
// call ruby script here
call('MY RUBY SCRIPT');
}
Don't hack into imagecache, remember every time you hack core/contrib modules god kills a kitten ;)
You should create a module that invokes hook_nodeapi, look at the api documentation to find the correct entry point for your script, nodeapi works on various different levels of the node process so you have to pick the correct one for you (it should become clear when you check the link out) http://api.drupal.org/api/function/hook_nodeapi
You won't be able to call the function you've shown because it is private so you'll have to find another route.
You could try and build the path up manually, you should be able to pull out the filename of the uploaded file and then append it to the directory structure, ugly but it should work. e.g.
If the uploaded file is called test123.jpg then it should be in /files/imagecache/thumbnails/test123/jpg (or something similar).
Hope it helps.
Users open files in our app through a QFileDialog. The order of the filenames is bizarre. What is determining the sorting order, and how can we make it sort by filenames, or otherwise impose our own sorting, perhaps giving it a pointer to our own comparison function?
The documentation and online forums haven't been helpful. Unless it's well hidden, there doesn't seem to be any sorting method, property, etc.
This is a primarily Linux app, but also runs on Macs. (I know nothing about Mac.)
Here is the juicy part of the source code:
QtFileDialog chooser(parent, caption, directory, filter);
/// QtFileDialog is our class derived from QFileDialog
chooser.setModal(true);
chooser.setAcceptMode(acceptMode);
chooser.setFileMode(fileMode);
QStringList hist = chooser.history();
chooser.setHistory(hist);
/* point "x" */
if(chooser.exec()) {
QStringList files = chooser.selectedFiles();
...blah blah blah...
From one of the answers, I tried an evil experiment, adding this ill-informed guesswork code at "point x":
QSortFilterProxyModel *sorter = new QSortFilterProxyModel();
sorter->sort(1); // ???
chooser.setProxyModel(sorter);
But this crashed spectacularly at a point about 33 subroutine calls deep from this level of code. I admit, even after reading the Qt4 documentation and sample code, I have no idea of the proper usage of QSortFilterProxyModel.
Are you using QFileDialog by calling exec()? If you are, you should have a button to switch the view to Detail View. This will give you some column headers that you can click on to sort the files. It should remember that mode the next time the dialog opens but you can force it by calling setViewMode(QFileDialog::Detail) before calling exec().
An alternative is to call the static function QFileDialog::getOpenFileName() which will open a file dialog that is native to the OS on which you are running. Your users may like the familiarity of this option better.
Update 1:
About sort order in screen cap from OP:
This screen capture is actually showing a sorted list. I don't know if the listing behaviour is originating from the Qt dialog or the underlying file system but I know Windows XP and later do it this way.
When sorting filenames with embedded numbers, any runs of consecutive digits are treated as a single number. With the more classic plain string sorting, files would be sorted like this:
A_A_10e0
A_A_9a05
Going character by character, the first 1 sorts before the 9.
.. But with numerical interpretation (as in Windows 7 at least), they are sorted as:
A_A_9a05
A_A_10e0
The 9 sorts before the 10.
So, the sorting you are seeing is alphabetical with numerical interpretation and not just straight character by character. Some deep digging may be required to see if that is Qt behaviour or OS behaviour and whether or not it can be configured.
Update 2:
The QSortFilterProxyModel will sort the strings alphabetically by default so there is not much work to using it to get the behavior you are looking for. Use the following code where you have "point x" in your example.. (you almost had it :)
QSortFilterProxyModel *sorter = new QSortFilterProxyModel();
sorter->setDynamicSortFilter(true); // This ensures the proxy will resort when the model changes
chooser.setProxyModel(sorter);
I think what you need to do is create a QSortFilterProxyModel which you then set in your QFileDialog with QFileDialog::setProxyModel(QAbstractProxyModel * proxyModel)
Here are some relevant links to the Qt 4.6 docs about it.
http://doc.trolltech.com/4.6/qfiledialog.html#setProxyModel
http://doc.trolltech.com/4.6/qsortfilterproxymodel.html#details
I don't think it depends upon the implementation of Qt libraries... But upon the Native OS implementation..
For example in Windows,
if you use QFileDialog, it will display the Files and Directories by Name sorted.. It is the same when used in other applications. In the sense that, if you try to open a file through MS- Word, it indeed displays the Files and directories as Name sorted by default..
And am not sure about other environments since am not used to them...
But in Windows, you can change the sorted order by right-click in the area of Files and Directories display and can select the options you like.. For e.g like Name,size,type, modified... And also which is similar, when you use an MS-Word application...
So, I believe it does depend on the Native OS implementation and not on QFileDialog's...