Specifying path to source files about malt HOT 15 CLOSED

Hi Rob,

two answers :

1. You first need to build your program or libraries with -g option to get the debug symbols in the binary so malt can find the source file paths.
2. If for some reasons, you built your program on another computer and install for example an RPM you might not have the sources in the same exact path when you run the MALT webview. In this case you can override some paths to redirect them to another directory by using the -o/--override option.

Eg, if you built with sources in /rpm/.... and run on another machine where you copied the sources in /home/me/

malt-webview --override /rpm/my-project:/home/me/my-project -i malt-123456-my-program.json

Fill free to precise if it is not working for you.

from malt.

I tried the -o option, but isn't working. Is there a way to display the original source path that web view is expecting?

from malt.

Ohhhh, thanks for the report, there is effectively a real bug.

You can see the path of the file in the webview source browser when it failed to fetch the file.

You are most probably in the case where your sources are pointed via a relative paths which trigger a bug in the webview making the './' disappearing while arriving on the mini-server.

I fixed it in master branch.

If you are in this case it might be fixed.

For relative path replacement you might use, or it might work automatically if you are located in the source directory.

malt-webview --override ./:/home/me/my-project -i malt-123456-my-program.json

If you can confirm it is your issue.

from malt.

from malt.

Pretty sur I still have an issue on the ./ in the target part of the redirection, I will give a look.

Waiting the patch, can you try to put a full path instead of ./ ?

from malt.

In principle I made the fix on master branch.
If you can check and confirm or infirm.

from malt.

from malt.

So if I well understand you get the issue since you moved to master branch and before you got success with probably the released version 1.2.1 ?

Or am I wrong ?

from malt.

from malt.

The issue related to fastSkip is related the the backtrace on which I skip the first entries not to look on my own internal functions. But the exact value to use might depend on the way your compiler as inlined some functions. In principle this is detected but I probably have an issue if you get the error.

You can try to see if another value of the skip parameter might fix your issue manually via the options :

malt -o stack:skip=3 ......
malt -o stack:skip=4 ......
malt -o stack:skip=5 ......

from malt.

from malt.

As you directly call cmakeyou can also try to build in release mode which enable all the optimizations which impact this. I apply it automatically if you use the ./configure script.

Probably i need to fix something in debug mode.

cmake .. -DCMAKE_BUILD_TYPE=Release .......

from malt.

from malt.

I moved the second issue your reported in #76 so I can close this one.

from malt.

from malt.