Navigating Latex Documents: Tips and Frustrations

[mrcleanhands]

Latex produces some really good looking documents and because of this I tried to pick it up. I was hoping it would make typing up my maths assignment a much quicker process.

What I find insanely frustrating is that because of all the marks up in a .tex file it becomes impossible to navigate and find where things are. Consequently, I spend an enormous amount of time trying to figure out what I need to edit.

It's selling point is you can concentrate on the content but I have found this not to be the case since your so busy finding out where the content actually is when you need to edit. I keep recompiling documents so I can actually check what I've typed out. Then I need to use this as a reference to find the original text in the .tex

Am I missing something? Is there no way to have some kind of preview up? How do you guys navigate massive documents?

 

[mfb]

Navigate to do what?

To read? I click on links in the table of contents.
Navigate to edit a chapter in general? I open the file for that chapter.
Navigate to edit something specific? I open the file for that chapter, and if there is no quicker way to reach it I search for 2-3 consecutive words via the search function.

 

[mrcleanhands]

hmmm maybe I should be splitting up my assignment question into different files then...

Is there an edit environment where I can have a main document and then click on the titled links to separate files?

 

[AlephZero]

I think the OP is talking about navigating around the "source" .tex file, not the output.

Use the same techniques as for any software large project. Put plenty of comments in the .tex files. Make the layout easy to read (you don't get extra credit for typing a math formula 5000 characters long all on one line and with no blanks). Split the input into logical manageable-sized pieces in separate files and use \include or \input to build the complete document.

You might like tools like http://mactex-wiki.tug.org/wiki/index.php/SyncTeX or similar, which let you click anywhere in the output PDF and take you to the corresponding place in the .tex file. (Well, that's the theory - but don't expect miracles).

 

[kith]

For formulas, you can also use an on-the-fly online renderer like http://www.codecogs.com/latex/eqneditor.php and copy the code to your file as soon as the output looks the way you like.

 

[TheoMcCloskey]

You might like tools like http://mactex-wiki.tug.org/wiki/index.php/SyncTeX or similar, which let you click anywhere in the output PDF and take you to the corresponding place in the .tex file. (Well, that's the theory - but don't expect miracles).

Texmaker has this feature built in and it can be quite usefull.

http://www.xm1math.net/texmaker/

 

[Solkar]

How do you guys navigate massive documents?

Using mnemonics for labels systematically.

 

[Hepth]

I use TeXMaker. I split chapters into separate Tex files, but this makes it a little tougher as you have to compile the whole thing for the preview.

But even without that, using chapter/section/subsection will make the structure appear in most good tex editors, and on the left you can just browse to it.

Also, when looking in the preview on the right, you can right click and "jump to tex" as well as the opposite in the code.

http://i41.tinypic.com/e82xpe.png

 

[mrcleanhands]

That left window with document structure is great - exactly what I'm looking for. It didn't work with my custom homework templates though as I'm using some other structure ( "problem" instead of chapter/section).

Also that jump to tex isn't working very well.

I think it's a software issue and I could probably work it out on my own.

If I sort that out then the only issue would be editing really big equations.

When I look at a whole bunch of equations thrown into a \begin{align} environment my eyes glaze and I have no idea what's going on. If I'm trying to edit something small in there it's a real pain!

How do you handle that?

 

[jackmell]

This is what I do in TeXnicCenter: While, I'm writing the paper, I build the source as DVI. That way I can use the YAP viewer and double-click on something in the document that takes me to the source code (approximately) where I clicked. Buy DVI won't show embedded graphics. But that's ok. Every once in a while, I'll build it as PDF to see how it actually looks with all the plots it it. But PDF won't take me to the source when I double-click. That combination of DVI for working edit, and PDF for checking every so often works good for me.

 

[Hepth]

The jump to PDF etc is always off, it usually puts you a bit after.
For long equations, remember it compiles out spaces and carriage returns so you can hit enter to make a new line in the code and it won't actually do anything.

This makes organization easier.
Also either label or comment every eq, it'll help when you're scanning for something ans makes referencing easier later.

 

[Solkar]

It's selling point is you can concentrate on the content but I have found this not to be the case since your so busy finding out where the content actually is when you need to edit.

I was not aware that someone said "LaTeX let's you focus on content", but I've learned general IT-"rule":
The more forceful and often is is claimed that a solution is business- (content-, data-) centric, the lesser that is the case.

I keep recompiling documents so I can actually check what I've typed out. Then I need to use this as a reference to find the original text in the .tex

My workflow is not so far away from that...

Am I missing something? Is there no way to have some kind of preview up? How do you guys navigate massive documents?

It's a good idea using mnemonics for labels, using an editor with a search history like vi, emacs etc, and to switch off automatic long line wrap.

Regards, Solkar