Bug 465880 - [bidi] JSDT UI needs a handler for comma-separated list of file path patterns
Summary: [bidi] JSDT UI needs a handler for comma-separated list of file path patterns
Status: NEW
Alias: None
Product: Equinox
Classification: Eclipse Project
Component: Components (show other bugs)
Version: 4.5.0 Mars   Edit
Hardware: All All
: P3 enhancement (vote)
Target Milestone: ---   Edit
Assignee: equinox.components-inbox CLA
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks: 464896
  Show dependency tree
 
Reported: 2015-04-30 02:21 EDT by Victor Rubezhny CLA
Modified: 2015-05-19 13:47 EDT (History)
7 users (show)

See Also:
markus.kell.r: review-


Attachments
The patch for Equinox Bidi Support required for Bug #464896 (9.75 KB, patch)
2015-04-30 02:25 EDT, Victor Rubezhny CLA
vrubezhny: review?
Details | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Victor Rubezhny CLA 2015-04-30 02:21:32 EDT
The handler should support the list of comma-separated file path patterns, like:

scripts/util/A*.js,scripts/util/,**/Test*

Such a handler is required for JSDT UI Include Path preference page translation for Mars.
Comment 1 Victor Rubezhny CLA 2015-04-30 02:25:43 EDT
Created attachment 252934 [details]
The patch for Equinox Bidi Support required for Bug #464896

The patch adds new handler for comma-separated list of file path patterns (that is placed in JSDT UI) to Equinox and updated an according JUnit test due to test the handler.

This handler required by JSDT UI in case if it is not allowed to be placed inside JSDT UI and contributed by an extension point to Equinox Bidi Support. See Bug #464896
Comment 2 Thomas Watson CLA 2015-04-30 08:43:00 EDT
Dani, this is an API enhancement request for Mars.  I am not in a good position to review the need or the fix on this one.  Markus are you a good person to review?
Comment 3 Markus Keller CLA 2015-05-04 08:44:07 EDT
I don't think a "file-patterns" processor should be released. The general problem of a delimiter-separated list of an existing StructuredTextType should be solved in a more general way that doesn't require a new handler for each SST/delimiter combination.
Comment 4 Tomer Mahlin CLA 2015-05-05 06:53:54 EDT
(In reply to Markus Keller from comment #3)
> I don't think a "file-patterns" processor should be released. The general
> problem of a delimiter-separated list of an existing StructuredTextType
> should be solved in a more general way that doesn't require a new handler
> for each SST/delimiter combination.

The only way for generic solution is to add a new signature of applyBidiProcessing API. There are structured text handlers (on the low level), which receive list of separators. We need to surface this configurable parameter in applyBidiProcessing, for example: 
  applyBidiProcessing(
       Control myCtrl, 
       String sttType, 
       String listOfSeparators)
We basically need to add a new constant for sttType (something like LIST_OF_SEPARATORS_BASED_SYNTAX) and create the structured text handler on the fly.
Alternatively we can pass a list of separators into existing API through sttType parameter (after all list of separators is a String as well): 
  applyBidiProcessing(
       Control myCtrl, 
       String listOfSeparators)
In the code of applyBidiProcessing we can check if listOfSeparators maps to any of known STT type. And if it doesn't we will interpret it as list of separators instead of STT type constant.
Comment 5 Samuel Wu CLA 2015-05-14 16:31:57 EDT
Any update on the fix proposed by Victor? It will be great if a target date can be provided so that WTP can plan accordingly. Thank you.
Comment 6 Markus Keller CLA 2015-05-19 13:47:39 EDT
(In reply to Samuel Wu from comment #5)
> Any update on the fix proposed by Victor? It will be great if a target date
> can be provided so that WTP can plan accordingly. Thank you.

Currently, there's no target date. It's too late for Mars.

And as said before, the org.eclipse.equinox.bidi framework will not add special cases like this. If someone wants to work on a generic solution, then please post it here, and then we can reassess.