I'd like to use docs generated by godoc to be served statically (on App Engine) for a project I created. I see that godoc starts a server and apparently the package docs are all generated dynamically. But is there a workaround to save the docs to the filesystem or what other options do I have?
If you're using GitHub, there's a pretty nice project for this already: http://gopkgdoc.appspot.com/
Documentation's at the bottom of the page. It's open source and with a little effort, you may be able to get it to generate documentation from non-GitHub services.
Related
As our RN application based on react-native#0.61.5, so we use expo#^0.37.0 to set up local development environment. But it seems that expo doesn't supply outdated docs, could only find recent version docs https://docs.expo.dev/versions/latest/.
Old docs often get lost, but thankfully there are a few ways of finding the old pages.
The first option is using the Wayback Machine from archive.org. The expo.dev site seems to go back to late 2020. A problem might be that some pages aren't crawled and might be missing.
And because it seems that their docs are open source with some commits going back three years. So you could in theory pull an old commit and build / run the docs locally.
You can use webarchive and old URL for documentation https://docs.expo.io.
In your case it is http://web.archive.org/web/20210305044934/https://docs.expo.io/versions/v37.0.0/ but not all pages are supported:
Use github online editor to check the outdate markdown docs is an optional solution.
Such as link: https://github.dev/expo/expo/blob/sdk-37/docs/pages/versions/unversioned/sdk/calendar.md#L1
I'm working on a project with ArcGIS/ESRI in WebStorm. According to the documentation, "Dojo is included with the ArcGIS JavaScript API so there is no need to host/reference/install it." This means WebStorm has no idea what require is:
How do I add support without installing Dojo/require?
Was just going to comment, but can't yet?! Anyway, I had to reengineer my web-site dev-ops to account for the ESRI javascript API. I am using 4.6 though...
The API, which includes dojo, may be installed, but it has to be configured into your site. Once you reference dojo/dojo.js, all will be good. But you need the config set-up first.
Start here:
https://blogs.esri.com/esri/arcgis/2015/11/13/using-bower-for-custom-builds-of-arcgis-api-for-javascript/
https://blogs.esri.com/esri/arcgis/2015/11/13/using-bower-for-custom-builds-of-arcgis-api-for-javascript/
There is a link to the API javascript resources. This will really help understand the process!
https://github.com/Esri/jsapi-resources
Also don't miss:
https://dojotoolkit.org/documentation/tutorials/1.10/modules_advanced/
https://dojotoolkit.org/documentation/tutorials/1.10/dojo_config/
http://dojotoolkit.org/documentation/tutorials/1.10/build/
I am just getting this stuff "almost" working. And actually had a build question to post and saw this. Hope it helps. I was using AMD and thought I had to go the requirejs route, but the dojo stuff worked out better for me. Check them both out and see what works for you! Good Luck!
I'm using gitlab to manage my repositories which has been quite pleasant so far. However the wiki could use a fair amount of improvement IMO.
It's really annoying when you already have documentation that you simply want to add to the wiki. Translating a table or pdf to gitlab flavored markdown can be time consuming and non-trivial.
My question is two-fold.
Has anyone had luck embedding files (specifically pdf's) into their github/gitlab wiki? This should definitely be possible on gitlab but is their an easy way to do it? I have a documentation repo set up, it would be awesome to embed files linked from the documentation repo.
If not embedding files into the pages. How do you efficiently translate a document to gitlab flavored markdown?
Had the same issue and landed here, maybe this helps somebody.
I didn't find a way to embed the .pdf into a page but what I did was to add the .pdf to the wiki folder (like a normal page). Initially this didn't show up in the sidebar nor was it searchable from Github.
It is possible however to link it directly from other pages or a custom sidebar (see for example osquery's wiki for a nice sidebar). Clicking on the link allows you to download the .pdf (didn't find a way to preview it in the browser)
It is possible that such action type is not allowed only in cloud gitlab version, cause gitlab-gollum lib documentation describes it so:
[[Gollum|gollum.pdf]]
I use Gitlab (self hosted) capabilities to render MarkDown pages regularly for documenting my projects.
The best way to copy paste html or pdf documentation and translate into perfectly formatted MarkDown is to use Typora. Typora has these very useful capabilities:
Support for Gitlab TOC rendering (i.e. [[_TOC_]])
copy and paste from HTML pages to MarkDown
upload of images
IMO is very easy to use for documentation with GitLab.
I really like the idea of this BundleTransformer plugin from NuGet, and I see that they have LESS support. But ...
After weeks of trying, I cannot find any decent example of how to use it. The documentation is unclear, to say the very least. This is the closest I could find, which does not compile or work.
who-could-ask-for-more-with-less-css
Well, the simple answer is that I could ask for more.
Can someone please assist me with the simple task of creating a bundle that has .less files in it, that will run? I am completely lost.
I would suggest reading the wiki page for bundle transformer and also, as the wiki page suggests, make sure you are familiar with ASP.NET bundling and minification.
If you follow the examples of usage you should be up and running.
I know this is an older post but figured I'd chime in. The Bundle Transformer Codeplex discussion board (Feb 2015) mentions that BundleTransformer is no longer considered a modern client-side build tool and they are recommending you go with Gulp or Grunt - which both have support starting in Visual Studio 2013.3 via VSIX extensions and more baked in to VS2014. FYI -This Scott Hanselman post explains how to get up and going with Grunt and Gulp in Visual Studio. I think if you're using BundleTransformer for simple Less compilation it works fine, but doesn't allow you to do things like generate CSS Sourcemaps which make your browser's dev tools aware of the Less source files, which in my humble opinion is critical for using a CSS preprocessor - the ability to see the LESS line numbers of your source files in the styles pane.
WHere can I get a XD version of dojo source like the one hosted on google? What I want to do is to host dojo source from my local CDN, and my custom dojo module in my web application. Is this a good practice? or I might as well just include the dojo source in my web app, and run the custom build?
Thanks,
You can build an xd version of dojo from the source code
Here are instructions on how to do it:
http://dojotoolkit.org/reference-guide/1.7/quickstart/custom-builds.html
See the section on "doing xdomain builds"
In our organization (a large one), we do have a CDN version of dojo deployed on internal CDN mainly since some of our webapps are not allowed to access extranet (firewall issues).
For performance, though, a custom build gives biggest boost since it is customized to the modules you need/use - once the custom build is done, you only need to ship a single compressed js output file and a small number of supporting files
When doing your custom build, you can use the xdDojoPath and loader=xdomain if you wish to use cross domain dojo to load your optimized js - see http://osdir.com/ml/cometd-users/2011-08/msg00050.html for some notes on this
Also see related SO question: Dojo on a CDN vs own install
The good news is that with Dojo 1.7+ and the new loader, you don't have to do anything special for a cross domain build (good answer above from #Vijay Agrawal, but I think that reference guide link may need some updating for 1.7) Just write your code to the new AMD format, use asynch:true, run the build tools to create layers, and deploy them on any server. AMD makes use of callbacks and many of the tricks the old Dojo xd builder used to employ, but in a much simpler way.
To support older code, there is a legacy cross domain mode mentioned in the loader docs.