path normalisation in "base"

[Daniel J. Luke]

one way to handle it could be:

% svn diff
Index: port.tcl
===================================================================
--- port.tcl	(revision 153771)
+++ port.tcl	(working copy)
@@ -2389,7 +2389,12 @@
         return 1
     }
     foreach filename $portlist {
-        set file [file normalize $filename]
+        # don't normalize absolute paths
+        if {"/" ne [string index $filename 0]} {
+            set file [file normalize $filename]
+        } else {
+            set file $filename
+        }
         if {[file exists $file] || ![catch {file type $file}]} {
             if {![file isdirectory $file] || [file type $file] eq "link"} {
                 set port [registry::file_registered $file]


Before -
% port provides `which nmap`
/Volumes/Data/opt/local/bin/nmap is not provided by a MacPorts port.
After -
% port provides `which nmap`
/opt/local/bin/nmap is provided by: nmap

> On Oct 11, 2016, at 9:07 AM, Daniel J. Luke <dluke at geeklair.net> wrote:
> see also:
> 
> https://lists.macosforge.org/pipermail/macports-dev/2014-May/026728.html
> 
> which I linked to the last time you brought this up.
> 
> (no one replied to that original post).
> 
> I expect no regular base contributor cares enough about that unsupported configuration to work on it - which means someone who does care (you, perhaps?) needs to generate and test a change that can be incorporated (otherwise we'll just keep having this conversation every year or so).
> 
> On Oct 11, 2016, at 4:43 AM, René J.V. Bertin <rjvbertin at gmail.com> wrote:
>> I'd like to understand a bit better why the base layer does path normalisation in a number of places where its use isn't immediately obvious to me, like for instance the action_provides procedure in the port script. If that's not so broad of a question that it cannot be answered with a single, succinct explanation.
>> 
>> I can see how it would probably be required in a sandboxing context, and I have no idea exactly what kind of sandboxing MacPorts does. (I do seem to recall whatever issues it had with e.g. a symlinked $prefix were resolved a while ago.)
>> 
>> To come back to action_provides: if the registry saves a port's "intended" paths (the ones stored in the software image tarball), why do a lookup of the actual/resolved path? That would make it impossible to check which port installs a symlink (to a file or directory installed by itself, some other port, or even to something in system space), regardless of whether there are "unexpected" symlinks in the path, no?

-- 
Daniel J. Luke