Hi all,
I would like to be able to
colorize, or otherwise distinguish, folded
text when it is folded. I find it hard to spot
in a program full of folding markers.
A folding marker “collapsed” of a completely
different color than the folding markers
“start” and “stop” might be sufficient...
TIA
All,
I've been playing with Textmate but haven't been able to do something
very simple (in my view) - write an Applescript that lets me get the
content of the frontmost Textmate document. This must be possible,
right?
If you could send me a pointer or a short example, that would be much
appreciated.
Thanks,
Thomaz
Hi, I just downloaded the experimental wiki bundle, and am getting
the following error:
Not Found
The requested page was not found.
Page: file://localhost(null)
Thanks.
All the best, Mark
_________________
Mark Eli Kalderon
Department of Philosophy
University College London
Gower Street
London WC1E 6BT
Dept webpage: http://www.ucl.ac.uk/philosophy
Personal wepage: http://www.kalderon.demon.co.uk
I recently acquired TextMate 1.5.3, and am overall very pleased with
it. However, I am running OS X 10.3.9, and several bundles do not work
correctly. In particular, several commands in the Latex and TODO
bundles do not work. Some of the problems were fixed by installing
Ruby 1.8 from Fink, but many others remain. Any help would be
appreciated.
Darryl Morrell
Hi all,
I would like to colorize the name
of the do and if loops in Fortran...
Those loops looks like this :
the_name:do i=1,10
some code here
…
enddo the_name
and similar for the if loops. I tried to
give a scope to those loop names but
with my poor knowledge of RegExp,
I cannot achieve something satisfactory...
Any help is much appreciated :)
Thanks in advance.
Hi,
As I've been using TextMate, I've noticed that some of the features
are, in my opinion, unnecessarily hard to discover.
Certain menu items are hidden from the user unless the user presses
the option modifier key while holding the menu open. Pressing option
causes one of the existing menu items to change meaning. For
example, the Edit menu of TextMate has a "Paste Previous" command
that converts to "Paste Next" when option is pressed. The Text menu
has the command "Execute Line Inserting Result" that converts to
"Execute Line and Replace With Result" when option is pressed. This
makes it harder to discover the available commands in TextMate
because the user must know to press option in order to see these
hidden menu items. As a recent Mac switcher, it was a couple of
months before I realized OS X menus had this hidden capability.
Besides making features harder to discover, hidden menus are harder
to select with only the mouse because the user must press "option" to
have the items available for clicking. Also, IIRC, there is
additional difficulty assigning keyboard shortcuts to these menu
items (I don't remember the details, but I thought there was
something about this on the TextMate mailing list).
On a related note, certain buttons on dialog boxes are hidden in a
similar way (for instance, the Find dialog's "Replace" button
converts to "Replace in Selection" when option is pressed). There
are no visible indications that the buttons are dual-purpose; indeed,
the user may never know they are there unless he stumbles over them
(as I did) by accidentally pressing option while the dialog box is open.
Would it be possible to avoid using hidden menus and buttons in
TextMate? Other than trying to keep the menu length a little shorter
or the dialog box a bit smaller, I've not come up with a benefit to
hiding these items from the user, and in my opinion there are several
benefits to making them visible all the time.
Thanks for your consideration,
Michael Henry
> On Sep 15, 2006, at 3:19 PM, Christopher Brewster wrote:
>
>> When I press apple button+R to typeset and view,
>> I get transferred to the Preview window but it does not show me
>> the current pdf
>> only the previous one.
>> Is there someway to correct this so that it refreshes the pdf it
>> is displaying?
>>
>
> Do you mean that it will not refresh the pdf after changes have
> happened to it and you have recompiled, or that if you compile a
> completely different file? If it is the former, then it is
> Preview.app's fault, and the only solutions I can think of would
> involve using AppleScript and slightly hacking Preview.app to allow
> AppleScript to drive it.
The former - it does not refresh the pdf after a recompile.
I do not know how to write AppleScript or hack Preview.app :(
> I would suggest using a different previewer, and in particular
> TeXniscope. Or you could use the TeXShop previewer, though that
> does not do pdfsync with TextMate (to my knowledge).
Yes I have tried TeXniscope several times.
2 problems:
- there is no search facility and this is important often to find
unresolved bibliography entries etc.
- it crashes frequently
This is why I have repeated returned to Preview.app. I have not tried
the TeXShop previewer.
Is anyone using Adobe Reader? Would that work better?
Christopher
First of all, one noob question: how do I make sure my mail is
recognized as a reply to an existing post in the mailing list? I have
tried clicking in the e-mail address on the mailing list page which
automatically generates an e-mail with the correct header … but
usually it isn't recognized as a reply to existing posts :-/
Thanks ;-)
> > (i) I can provide you with a list of `official' LaTeX commands
if needed.
> >
> > Obviously, something similar should be implemented in math mode as
> > well (start with dark red which changes to red once a command is
> > recognized).
>
> allan: (i) I prefer that as well — if we can get a list, that’d
be
> great!
>
> me: I agree too. Max (and anyone else as well), if you provide us
> with the commands, for within math and out of math, then we'll add
> them in.
Ok, I will compile a list over the course of the week. Basically, I
will start by copying the standard commands from the various handbooks.
Please document how to add your own commands to that list! (E. g. I
have created commands like \norm, \scpro (scalar product), \ket,
\bra, \C, \R, etc.)
> allan: (ii) this is because $ is a smart-typing pair, and it finds
> that there is a “starting” $ outside the screen — not sure how
> to best address that (probably we want to keep it a smart typing
pair)
Definitely, definitely, definitely keep smart-pairing $s, it was my
#1 syntax error with TeXShop and very tedious sometimes to find the
erroneous line of code.
> > (iii) Footnotes and marginpars should be marked within the text;
> > perhaps the footnote's/marginpar's text could even be folded (not
> > sure if automatic foldmark recognition works on a bundle level).
>
> allan: (iii) marking: yes, folding: we can’t do w/o having the
> braces align
>
> me: The text in the footnote gets a scope of meta.footnote.latex
> So if you add that scope to your coloring theme and add some color to
> it, you'll get highlighted footnotes. Not immediately possible for
> marginpars, but we are thinking of matching those as well and giving
> them some scope extending both footnotes and marginpars, so that you
> could color both of them at the same time.
Yes, it is immediately possible. I've just added marginpar to the
footnote environment (so its scope is now both footnotes and
marginpars alike), took about 10 seconds :-)
I already figured folding would be a more intricate matter.
> Update: marginpars are now matched with scope
> meta.paragraph.marginpar.latex, so you can color them as you like.
>
> > (iv) Use input instead of include when dragging a .tex document
> > into another: afaik include is deprecated and input is preferred
> > instead. You cannot `include' some bundles for instance (diagxy
> > comes to mind), you have to use input instead.
>
> allan: (iv) probably then we should even markup include as
> invalid.deprecated
>
> me: I don't think \include is deprecated. I use \include for
> different chapter of a book etc, and it does a number of special
> things like clear all the floats, start on a new page, generate
> separate aux files etc.
input also creates separate aux files. I just ran into some problems
with include and have used input ever since. Page breaks are not a
concern since I usually have separate files for each chapter -- the
chapter command takes care of the page break.
> Of course, you can easily change this in your own copy, by opening
> the bundle editor and finding the appropriate drag command. All you
> have to do is change the \\\\include that appears there to \\\\input.
Done.
> > (v) Forget about the deprecated math environments `\[ … \]' and
`\
> > ( … \)', just remove them from your bundle. Guidelines by
relevant
> > journals suggest to use specific environments anyway (align for
> > instance).
>
> allan: (v) I’m all for removing stuff, especially when
deprecated ;)
>
> me: I would need some stronger evidence that it is deprecated. As far
> as I understand, the alternative to \( \) would be $ $, and I don't
> see what advantages that has except that it is much harder to parse
> the file looking for math in it, and if you miss one of the $ you
> might not find out until much much later.
>
> As for \[, \], I again don't think they are deprecated, but I would
> love to be proven wrong. The only alternative I see is \begin
> {equation*}, which requires the amsmath package. I might possibly
> agree with you that in the context of writing math papers for
> submission to journals, one might want to avoid \[, \], (though still
> I would need to see some strong evidence for that), but I don't see
> why I shouldn't be using it in the notes for my students for
> instance. Replacing them all by \begin{equation*} is a single search
> and replace anyway. (Note to self: Actually, having a command that
> toggles the various math environments might not be a bad idea at
all.)
I usually use align and align*. I have never personally used \[ … \]
and I have seen only one person use it, and that person still writes
his (brilliant) papers in plain TeX.
It doesn't really bother me if you keep it, but I just thought that
we could very well eliminate everything which is superfluous in a way.
> Do you have references to these guidelines of these journals
> (including non-math/physics journals)?
One of the guidelines I use is the revtex guidelines which is used
for a wide number of journals, including Phys. Rev. A-E and Phys.
Rev. Lett.
> > (vi) Inline formulas should have a grayish background so it's
> > easier to see where formulas begin and where they end. (This one is
> > also important to me.)
>
> allan: (vi) that’d be a theme-thing, as long as the grammar marks
> them up
>
> me: Do you want the background for inline formulas to be different
> than the one for multiline formulas? Because as things are already,
> you can target the scope math.tex in your theme, and that would color
> all math however you tell it to.
Yes, I want it to be different. I add a commented line above and
below the align environment, i. e.
%
\begin{align}
formulas
\end{align}
%
So I don't need any background. In either case, it is desirable to
have different settings for inline formulas and displayed equations.
> > (vii) You already have tab triggers for section, subsection and
> > subsubsection, so how come chapter is missing?
>
> allan: (vii) oversight (and we rarely need it)
>
> me: Basically, initially we were trying to keep the number of
> snippets at a bare minimum because of the way they were showing up in
> the menu, as a flat list. On top of that, most of those can instead
> be accomplished via the “Insert Command…” command, along with
> customizing it through the LaTeX Configuration file, so the other
> maintainers had a hard time even convincing me to keep the section
> ones in. Now that we can create submenus, we've toyed with the idea
> of adding more snippets, so any ideas on what other snippets to
> include are welcome.
I know I can add it. However that is something I definitely think is
missing. \chapter is one of the most basic commands in TeX and I
would strongly suggest to add it.
> > (viii) A way to execute bibtex (and pdflatex twice to see whether
> > all additional citations have been included).
>
> allan: (viii) latexmk.pl should do that — maybe we should make this
> the default, not sure if that would bother anyone (maybe some
> workflows would mean much more time typesetting)
>
> me: latexmk.pl will actually do a lot more for you. The Help file
> should describe how to set it up. I'd personally prefer not to have
> it as the default. It is however easy to set it in your system.
I think I wasn't expressing myself clearly here: when you run bibtex
via the LaTeX bundle, it does just that. However, I think it is a lot
more useful if you actually pdflatex the document twice and display
the document so you can check whether or not it worked.
> > (ix) Closing environments: when I manually type \begin{environment}
> > and then close the environment, the `\end{environment}' which is
> > added is indented like the text within the environment. Hence,
> > TextMate's code folding does not recognize the block.
>
> allan: (ix) http://macromates.com/ticket/show?ticket_id=B34CCC0C --
> the request might be granted, but use begin⇥ or ⌘{ until then
>
> me: just follow the workaround suggested in that ticket (and most
> importantly, don't manually type \begin{environment}).
No, I usually don't. But sometimes I need to break a displayed
formula into two, I use that command.
> > Now concerning the Help. The help is nicely structured, although I
> > miss a nice webpage with the key features of your bundle. That
> > would have helped me to use more of the functions included in your
> > bundle.
>
> Hm, that was actually partly our intent with rewriting the help file
> this way. To make it easy to find out how to do stuff. Looking at the
> bundle is of course the best way to find out what commands are
> available. How is the help lacking in showing you what the bundle can
> do?
> >
> > (i) A glossary of TM_LATEX_BLABLA variables. That would be really
> > helpful, especially for people who just want to check out what you
> > can do with the LaTeX bundle.
>
> There are basically only five such variables, most of them having to
> do with more particular workflow setups:
>
> TM_LATEX_VIEWER if you don't want to use the built in previewer
> TM_LATEX_ERRLVL if you want some finer control on what errors
> show up
> TM_LATEX_COMPILER if you want to use latexmk.pl
> TM_LATEX_OPTIONS for any options you might want to add to the
> command line call to the compiler.
>
> and finally, TM_LATEX_MASTER, the only one of more frequent use, when
> setting that a master document should be used for the compiling.
That needs to be documented in the form of a howto.
> > (ii) Overview over key functions (auto completion of citations,
etc.).
>
> Isn't the explanation in section 5.2 of the manual adequate for that?
> In general that's the purpose of the entire section 5. In what ways
> is that failing?
In the way that people like me don't read the help until section 5.2
to discover new features. I know I sometimes should, but it's just
the way it is. In this way, I would suggest to add one section in the
very beginning (section 1.2 or so) about basic features. If they are
hidden in section 5.2, fewer people will actually use them -- which
is a pity.
I have given the help some more thought and I think that also one
chapter about customizing the LaTeX bundle is missing. Somehow I
haven't come across a good documentation on how to edit bundles,
something like a HowTo is definitely missing.
For me, the most helpful kind of documentation is one that explains
by example (e. g. Samba by Example). So I would suggest to write
HowTo sections on `Getting Started', `Big LaTeX Projects',
`Customizing The LaTeX Bundle'. I would be willing to make additions
of my own.
Ok, that's all for now.
Max
Hi,
I am slowly adapting to the language defs. Great.
Now, I am having difficulties to specify recursive blocks for a
language.
So far, I got it working pretty good for things like
-----------------
%function foo() {
%{
%}
%}
-----------------
as I created a matcher for {%..%} that calls itself and thus the matcher
for the function which for simplicity I state here as
begin = '%function .*\(\) \{'
end = '%}'
will match the last %} and not the previous one. Cool.
Now, how do I allow for this case:
---------------
%function foo()
{
%{
%}
%}
---------------
I can't get it to work even though I know that the matcher only
matches whole
lines. The language I try to define is an obsolete old IBM mainframe
macro
language called Net.DATA (please, don't ask why).
So a block starts either as '%keyword_and_more {' or as '%
keyword_and_more\n{'
and always ends in '%}'.
Is it possible to define this? I assume that I need to define two
rule sets to
simulate the two states and then either start in state A or B and do the
recursion within each set. Right?
Andreas Pardeike
PS: If you like, you can see my current definition:
{ scopeName = 'source.netdata';
fileTypes = ( '' );
foldingStartMarker = '%.*\{|%(?i:if)';
foldingStopMarker = '%\}|%(?i:endif)';
patterns = (
{ include = '#embeddedstuff'; },
{ name = 'source.netdata.comment';
match = '^%\{.*?%\}';
},
);
repository =
{ embeddedstuff = { patterns = (
{ include = '#shtml'; },
{ include = '#netdata'; },
);
};
netdata = { patterns = (
{ name = 'keyword.control.netdata.conditional';
match = '%(?i:if|else|endif)';
},
{ name = 'keyword.control.netdata.include';
match = '%(?i:include).*';
},
{ name = 'keyword.control.shtml.include';
match = '<!--#%(?i:include).*-->';
},
{ name = 'source.netdata.block';
begin = '(%\{(?i:macro)?)';
end = '(%\})';
captures = { 1 = { name = 'keyword.control.netdata'; }; };
patterns = ( { include = '#embeddedstuff'; } );
},
{ name = 'source.netdata.function';
begin = '(%function\s*\(([^)]*)\))\s+(\w+\(.*\))\s*(\{)';
end = '(%\})';
captures =
{ 1 = { name = 'keyword.control.netdata'; };
2 = { name = 'source.netdata.function.type'; };
3 = { name = 'source.netdata.function.name'; };
};
patterns = ( { include = '#embeddedstuff'; } );
},
{ name = 'source.netdata.block';
begin = '(%.*?)\s*\(.*\)\s*(\{)';
end = '(%\})';
captures =
{ 1 = { name = 'keyword.control.netdata'; };
2 = { name = 'keyword.control.netdata'; };
};
patterns = ( { include = '#embeddedstuff'; } );
},
{ name = 'source.netdata.block';
begin = '(%.*)\s?(\{)';
end = '(%\})';
captures =
{ 1 = { name = 'keyword.control.netdata'; };
2 = { name = 'keyword.control.netdata'; };
};
patterns = ( { include = '#embeddedstuff'; } );
},
);
};
shtml =
{ name = 'source.netdata.embedded.shtml';
begin = '(?=<!--#)';
end = '-->';
patterns = (
{ name = 'keyword.control.shtml.include';
match = '(?i:include.*?=".*?")';
}
);
};
};
}
Is the ability to make templates broken again?
I just tried to add another template and I cannot get any new ones to
work. All existing templates, including the ones I made work, but any
new ones will not load. I even tried copying an existing template and
the copy will not load.
Thanks
Robert
When I press apple button+R to typeset and view,
I get transferred to the Preview window but it does not show me the
current pdf
only the previous one.
Is there someway to correct this so that it refreshes the pdf it is
displaying?
Thanks,
Christopher
*****************************************************
Natural Language Processing Group,
Department of Computer Science, University of Sheffield
When doing multiple finds, the textmate window goes a light shade of blue,
and then stays that way unless I search backwards. How do I turn this off,
and what is it supposed to mean? I've looked in the manual, and don't see
anything about this (not under Find).
Thanks,
Bob Sidebotham
Hi,
my glacial migration to ruby continues: I've been dipping into the
TextMate book and noticed that ^H should bring up doc refs for ruby
keywords. Unfortunately using the command (in build 1252) just threw
up a tooltip complaining about "url_encode" being an undefined
method. I think that the file web_preview.rb in
TextMate.app/Contents/SharedSupport/Support/lib
needs to have the following pasted in after "require erb" :
include ERB::Util
Well, that fixed the problem for me.
Cheers,
Paul
1. Type StringScanner into a Ruby file. The full documentation
appears, with links to methods.
2. Click one of the method links. The documentation for the method
appears, with a link at the top back to StringScanner.
3. Click the StringScanner link.
Expected: The same text and markup will appear it did in step 1.
Actual: The list of methods at the end of the page is truncated.
-- F
Greetings,
I believe there is a problem in the OCaml syntax highlighting. It has
to do with:
let xxx = 10
and yyy = 20
If you want it to color that correctly, you have to:
let xxx = 10 and
yyy = 20
Can someone verify? I have posted two screen shots:
http://lifewithchrist.org/OCaml_tm.png <- wrong way
http://lifewithchrist.org/OCaml_tm_correct.png <- right way
Notice on the first image, how the vars turn italic and purple. But
also notice how the try statement below does the exact same thing.
Jeremy
I was at a local Mac Users Group meeting last night and during the Q&A
session, they were writing down all Q&A's using SubEthaEdit. The one
cool thing was there was 2-3 people writing this stuff down, populating
URLs, etc.
I read Allan's "Future" posting, but I saw nothing about it and was
wondering if TextMate would ever get this cool feature? It's probably
the only thing I'm left wanting at the end of the day.
Thanks!
Derek
I would like to be able to use Textmate with very large directory trees. I
have a tree that has well over 100,000 files in it, and when I try to open
the top-level directory, I get the spinning beach ball. I end up having to
terminate Textmate, losing any other open sessions.
The behavior I'd like is for Textmate to only look at the top level
directory, and to browse into lower level directories as I open them. I
*think* that it is actually trying to read in the entire directory tree, but
this is not a good idea in this case (my file system also happens to be NFS
mounted).
Any chance of supporting this in the future?
Has anyone else run into this?
Thanks,
Bob
I'm using a perl search/replace string to re-format documents into
HTML. I'm using the following for regular paragraphs:
s/^[^\n\t\<].*/<p>$&<\/p>/g;
That takes care of single-line paragraphs like this:
Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do
eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim
ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut
aliquip ex ea commodo consequat.
This is just like what you'd get from Markdown, paragraphs separated
by a blank line are wrapped in a paragraph tag. But I'd like to add a
rule that looks for paragraphs that have hard-breaks in them, like this:
Lorem ipsum dolor sit amet,
consectetur adipisicing elit, sed do
eiusmod tempor incididunt ut labore
et dolore magna aliqua.
and wrap them in markup like this:
<p>Lorem ipsum dolor sit amet, <br />
consectetur adipisicing elit, sed do <br />
eiusmod tempor incididunt ut labore <br />
et dolore magna aliqua.</p>
The key is finding lines that end only to be followed by more lines
in the same paragraph. My RegEx-fu is okay, but not great, so I've
come close, but I can't get it right.
Thanks in advance.
Subversion bundle, Commit window.
I wrote a long, long commit message. I decided I wanted to refer to
an earlier message (so I'd remember what changes happened since the
last commit). I choose a "Previous Summary" from the menu. The
message-text area is filled with the historical text.
Expected: I should be able to recover the message I was working on,
via a control or at least via Undo.
Actual: My work on the message is destroyed.
-- F
Hi,
Has anyone tried this intel build of texniscope?
http://www.akropolix.net/rik0/blogs/2006/04/10/texniscope-intel/
I have tried it and everything seems fine, apart from pdfsync (it can
link from and to the latex document ok but just not as precise as the
original powerpc build). just wondering if anyone else have had any
experience with this build and knows how to make pdfsync to work
properly.
thanks,
Jeff
> Yes that should be doable, at least in the case where the (fold),
> (end) pair is used, once we capture that case in the grammar. The
> reason it works with environments is because I can set the input to
> the current scope, so provided each entire section gets its own
> scope, it should be doable.
> Similarly, the Change Environment command could be made to work for
> sections as well, allowing you to quickly change a section to, say,
> a subsection, with automatic renaming of the label prefix as well
> as the % section (end) part.
I would also suggest that you make the fold marks specific to that
part, i. e. (end section) or something like that.
Not only does it make the code more readable, but I suppose it's also
easier to filter and look for the right (end) …
Max
I tried to track down the other files.
I could not find them until I read about using 'find in project'.
They were inside a directory with the suffix .texp which was
generated it appears by texshop.
The files were in a 'soft links' subfolder.
Deleting this directory (which must have been generated in the past)
solved my duplicate entry problem.
Thanks,
Christopher Brewster
is there some change to the shift+control+w ?
before i could invoke that, type, hit tab and hit enter to get
<div>
|
</div>
now i get
<div>
|</div>
and i dislike that quite a bit, has some change been going on behind
the scenes?
______________________
Andreas Wahlin
Webbutvecklare
Webbhuset AB
Östra Hamngatan 45, 411 10 Göteborg
Telefon: 031-339 19 19, Direkt: 031-739 18 20, Fax: 031-711 12 20
www.webbhuset.se
Privileged/confidential information may be contained in this message.
If you are not the addressee indicated in this message (or
responsible for delivery of the message to such person), you may not
copy or deliver this message to anyone. In such case, you should
destroy this message and kindly notify the sender by a reply email.
Thank you.
Hi,
can we get the same end-of-text-click behavior as in Xcode, TextEdit
and BBedit?
I am talking about this:
(1) If you click in the empty area at the end of a text document, the
text cursor
jumps either into the last line exactly vertical to the clicked
position
OR
(2) The cursor jumps at the very end of the text.
Most apps in MacOS X do it the (2) way. TextMate does it the (1) way.
It is simply
annoying.
Just a comment for an otherwise great app,
Andreas Pardeike
I've been playing around with running python scripts from TextMate
(PyMate) which is great. Is there a way to pass arguments or input to the
script. I realize I could just as easily run this from the command line,
but I figured I'd check anyway.
