isra00 / online-phpdoc Goto Github PK

...the same way that Travis-CI does with console output: https://travis-ci.org/isra00/class-triggers

Start/stop tracking repos with http://codepen.io/craftedpixelz/pen/qFEny

If the design is great, this will really make a difference in the whole product and its adoption.

Hi there
Once i have authorised the app i get taken to the start page as expected. But my name and my repos are not showing.

Like Travis does:

When I try to login I get the notification that the app wants read and update access; is the update necessary?

Instead of rendering source code in the docs, simply link each function to its line in GitHub.

http://emberjs.com/api/

Instead of copying all files and modifying them, just extend from Responsive template and override the desired files.

When implementing the planned feature multiple versions we can benefit from this: http://phpdoc.org/docs/latest/for-developers/specifications/documenting-multiple-versions.html

(like Travis does). It will improve user engagement.

Patrones:
http://behance.vo.llnwd.net/profiles/51187/projects/143632/511871225820562.jpg
http://behance.vo.llnwd.net/profiles/51187/projects/143632/511871225822719.jpg

Más:
http://www.behance.net/Gallery/NPM-Pattern/144115
http://www.behance.net/Gallery/TIME-FOR-BREAK/143632

http://www.phpdoc.org/docs/latest/for-users/introduction/definitions.html#term-abstract-syntax-tree-ast

It will allow smoother transition to Symfony... but it's worty the effort to implement Composer in the prototype?

table log_job (engine=ARCHIVE)

• id_job
• started
• time_download
• time_generate
• generator_output
• generator_status

Advantages:

• Optimize DB by making the job table lighter (move the current output, time_download and time_generated fields) and having a log table with the ARCHIVE engine
• Separate the job queueing from the logging aspect, which allows a smoother transition to an external queue system (RabbitMQ, etc) and optimizes future log analysis.

Use case
1.- User generates docs for a given GitHub repo
2.- User deletes the repo from GitHub
3.- Repo code and docs stay in the server forever.

We should write a script which checks that every repo is still existing in GitHub.

Possible implementations (they're not supplementary but complementary):
1.- Removal webhook (if it exists!)
2.- Cronjob. We can optimize it checking only the repos with no activity in the last months.

Is it possible to use this on private repositories?

As a user, i must be able to:
1.- I start tracking a repo. For doing so, the system will add a webhook to the repo with the notification URL
2.- I stop tracking the repo. The webhook from GitHub and the rows from repo and job tables are removed.

They should include a custom var specifying the repo name and/or ID

http://symfony.com/doc/2.0/components/process.html

http://github-phpdoc.israelviana.es/gh-auth.php?error=access_denied

"It is able to manage versions of your code to generate documentation for all of them in a single tree (without the overhead of re-parsing everything for each version of course)."

http://fabien.potencier.org/article/63/sami-yet-another-php-api-documentation-generator

How long does it take until the Repo gets crawled when it's in the status "Waiting in the queue" (and I synched new commits to it).

Another question: do you specifically watch for code file changes or does a sync of GH files (i.e. Readme.md) also trigger the online-phpdoc parser?

Use new time_download and time_generate fields in job table. Update create_tables.sql with those fields.