I'm getting the following error when I try to run the TODO Bundle
"Show TODO List" I can't figure out what it means.
/opt/local/lib/ruby/1.8/erb.rb:354:in `trim_line2': private method
`scan' called for #(NoMethodError) from /opt/local/lib/ruby/1.8/erb.rb:
311:in `call' from /opt/local/lib/ruby/1.8/erb.rb:311:in `scan' from /
opt/local/lib/ruby/1.8/erb.rb:534:in `compile' from /opt/local/lib/
ruby/1.8/erb.rb:702:in `initialize' from /Applications/TextMate.app/
Contents/SharedSupport/Bundles/TODO.tmbundle/Support/todo.rb:52:in
`new' from /Applications/TextMate.app/Contents/SharedSupport/Bundles/
TODO.tmbundle/Support/todo.rb:52
I'm running a macports version of ruby (ruby 1.8.7 (2008-08-11
patchlevel 72) [i686-darwin9])
Any help you could provide would be greatly appreciated.
Tony
The new way of selecting snippets with "command + #" (when for
instance there's more than one assigned to a tab trigger) forcing you
to use "command" key instead of just number is really annoying... to me.
Is it possible to cancel this new behavior somehow?
And if not, what's the reason behind the change? I might be better
able to get used to it if I can understand the benefits. ;)
Thanks!
Hi all,
Below is a patch for the Ant bundle. It fixes up the htmlization
of the Ant output (escapes the output correctly so that, for example,
generics display correctly). It also turns htmlization back on, so
that you can click on error messages to jump to the appropriate line
again. Finally, it adds source.java to the scope of the build command
so you can build from a source file, not just from the build file.
I haven't turned column parsing back on... yet.
Be well,
Will :-}
Index: Commands/Build.tmCommand
===================================================================
--- Commands/Build.tmCommand (revision 11139)
+++ Commands/Build.tmCommand (working copy)
@@ -13,7 +13,7 @@
TM_PROJECT_DIR=`dirname "$TM_PROJECT_FILEPATH"`;
-#TM_ANT_BUILD_PARSER=htmlize_ant.py
+TM_ANT_BUILD_PARSER=htmlize_ant.py
if [[ "$TM_ANT_BUILD_PARSER" == "" ]]; then
TM_ANT_BUILD_PARSER=pre
@@ -42,7 +42,7 @@
<key>output</key>
<string>showAsHTML</string>
<key>scope</key>
- <string>text.xml.ant</string>
+ <string>source.java, text.xml.ant</string>
<key>uuid</key>
<string>C7888301-C304-4EF2-A15F-5681A7D6D13C</string>
</dict>
Index: Support/bin/htmlize_ant.py
===================================================================
--- Support/bin/htmlize_ant.py (revision 11139)
+++ Support/bin/htmlize_ant.py (working copy)
@@ -1,6 +1,6 @@
#!/usr/bin/python
-import sys, os.path, re, os
+import sys, os.path, re, os, cgi
matcher = re.compile(
r'(/.*?):(\d+):\s*(.*)$'
@@ -20,6 +20,8 @@
sys.stdout.flush()
+print "<pre>"
+
## read all data from stdin
lastLine = None
line = sys.stdin.readline()
@@ -33,15 +35,13 @@
match = matcher.search(line)
- print "<pre>"
-
if not match:
- print line
+ print cgi.escape(line)
else:
fn = match.group(1)
if proj_dir and fn.startswith(proj_dir):
- short_name = fn[len(proj_dir):]
+ short_name = '/.../' + fn[len(proj_dir):]
else:
short_name = fn
@@ -56,19 +56,21 @@
## # print lastLine[(brktInd + 2):], colInd
- print line[:match.start()].rstrip(),
- print '<a href="txmt://open?url=file://%s&line=%s&column=%d">
%s:%s: %s</a>' % (
- fn, match.group(2), colInd, short_name, match.group(2),
match.group(3)
+ print cgi.escape(line[:match.start()].rstrip()),
+ print '<a href="txmt://open?url=file://%s&line=%s&column=%d">
%s:%s</a>: %s' % (
+ cgi.escape(fn), match.group(2), colInd,
cgi.escape(short_name),
+ match.group(2), cgi.escape(match.group(3))
),
- print line[match.end():]
+ print cgi.escape(line[match.end():])
- print "</pre>"
sys.stdout.flush()
## read next line
lastLine = line
line = sys.stdin.readline()
+print "</pre>"
+
print """
</body>
</html>
Hello there,
On 1.5.8 r1496 the LaTeX bundle seems to ignore the TEXINPUTS variable.
I turned on the debug output:
===
%!TEX Directives: {}
[...]
TEX package list = ['indentfirst', 'setspace', 'pdfsync', 'amsmath',
'amssymb', 'amsthm', 'fancyhdr', 'lastpage', 'extramarks', 'chngpage',
'url', 'graphicx', 'graphviz']
texMateVersion = $Rev: 10791 $
engine = pdflatex
texCommand = latex
viewer = Skim
texinputs = .::/Applications/TextMate.app/Contents/SharedSupport/Bundles/Latex.tmbundle/Support/tex//
fileName = A2.tex
[...]
===
Python seems to be able to pull the env variable properly:
===
toirse:~ thatha$ python
Python 2.5.1 (r251:54863, Jul 23 2008, 11:00:16)
[GCC 4.0.1 (Apple Inc. build 5465)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import os
>>> print os.getenv('TEXINPUTS')
/Users/thatha/.latex//:
===
Any ideas what's wrong?
I recently transitioned to OS X from Linux and I'm wondering if
setting the variable in .profile is enough?
Thanks,
Ian.
I'm getting this error message when running the command tidy from the
XML bundle on any xml doc:
Errors: dyld: lazy symbol binding failed: Symbol not found: _gzdopen
Referenced from: /opt/local/lib/libxml2.2.dylib
Expected in: dynamic lookup
Here's a couple of questions:
1) How can I get access to the error above which appears in a yellow
fading window in a location I can copy and paste.
2) How is the default path set for running textmate commands from
bundles -- perhaps I can remove /opt from the path?
3) Anybody know what this error message means?
Thanks
Hi,
Following from an different email thread earlier this year. Link:
http://www.nabble.com/Searching-for-Objective-C-documentation-td16722396.ht…
As some of you know, Textmate has this very good script for performing
correctly the searches on Apple Docset with the /Developer/usr/bin/
docsetutil tool.
The command is bound to Ctrl - h, and will also search the unix man
(3) pages.
Offers a multiple - selection popup.
To access this feature you must type in and select the word into your
document before proceeding. However it strikes me that adding an
alternative method to use the script (probably a Search box) could be
useful.
There;s nothing wrong with the Ctrl - h shortcut just that its rather
cumbersome if you are searching
speculatively for a series of related classes. For that we really can
prefer an open Search box where
the choices are typed in, without affecting / editing the current open
document.
Basically any improvement of the current functionality thats going to
let me stay inside TextMate and
not lead me to an external program or tool such as AppKiDo. Actually
there also used to be QuickSilver + Developer Documentation Plugin,
but that was broken since the Docset format came in.
Perhaps something like the "Select Bundle Item" window (Ctrl-Cmd-T),
which can be toggled visible / invisible and left open for multiple
searches would be the best way to go.
My ruby is - 0/0 unfortunately so cannot put up any code myself for
what i am trying to propose here. But anyway the relevant place to go
is file docset_query.rb
There could be an obstacle because how should we handle multiple
results?
Of course already there is a place already in the Ctrl-H cmd where the
mini selection list pop-up is shown.
http://svn.textmate.org/trunk/Bundles/Objective-C.tmbundle/Support/lib/docs…
Function "def show_document (results, query)"
If we write a new command in this file and choose not to call this
above function, then its going to mean that we can use other existing
function definition in the ruby script and we can bypass the showing
of the results mini- pop-up window, right ?
Or perhaps we will need some kind of re-write because there another
place at the beginning of the file where "the selected word" is used
for input? Would we want a different input going into a new ruby
function. The user input is probably going to be from a search box.
And later we must print any multiple search results (called 'results'
array). Printed back into e.g. an NSTableview of the bottom half of
our search box.
I cannot find any existing feature improvement / request for this yet,
but still looking.
Cheers
Gulliver
dreamcat7
dreamcat7(a)googlemail.com
I'm thinking to create a new bundle to help people translating documents.
I think the most important thing is to have the original text handy, so
here's my idea.
#begin
#it
Questa è una parte di testo in italiano
#endit
#en
This is a parte of text in italian
#enden
#end
-- now I just have some code, which doesn't change.
[x | x <- [1..100]]
So we have markers for the text and markers for the languages.
In this way we can fold whenever we already finished to translate something.
After there will be a command that looks how many languages there
are and generates n files in from <filename<language>.ext>.
And maybe other commands to look for on the internet for words.
What do you think? Could it work??
It' s not as good as using something more xmlish but I think that for
articles or just normal text it could be fine...
--
View this message in context: http://www.nabble.com/Translations-tp21525170p21525170.html
Sent from the textmate users mailing list archive at Nabble.com.
Hey,
the phpDoc stuff in the PHP bundle always inserts an @version tag with
value $Id$ when I insert doc_h. According to the phpDocumentor docs (http://manual.phpdoc.org/HTMLframesConverter/default/phpDocumentor/tutorial…
), this is for integration with SCM ($Id$ to be substituted with the
last-change revision for the given file), but how do I automate this
within TextMate, or does this belong in a post-commit hook?
Thanks,
Ben
Hi, I'm fairly new to TextMate but I've been through the mailing list
archives and was unable to determine if the following is possible.
In general, I'd like TextMate to display a unicode character in place
of a sequence of ASCII characters. For instance, I'd like to be able
to type => but have the editor display ⇒ (U+21D2).
The saved file will need to be ASCII so I'm not looking for a way to
replace characters, just a way to choose a different representation.
Is there a simple way to do this?
Thanks,
Jonathan
--
ten ni hibari ningen umi ni asobu hi zo
larks in the sky
people in the sea
a holiday
Hey everyone,
Recent TextMate convert here and boy do I love it - trial expiring
within the week, so will be a paying customer from then on. ;)
Have a question, though, with regards to sorting in the Project Drawer.
Is it possible to have entries sorted by type, and only then by
name? I'd like to keep directories together. (At the top,
preferably. ;)
I found the following reference (and its references) in the mailing
list archives:
http://comox.textdrive.com/pipermail/textmate/2004-December/001971.html
This message seems to imply the sorting is hard-coded.
Alexander Deruwe
