1 |
On Mon, 21 Sep 2015 09:09:04 -0700 |
2 |
Zac Medico <zmedico@g.o> wrote: |
3 |
|
4 |
> On 09/20/2015 10:39 PM, Ulrich Mueller wrote: |
5 |
> >>>>>> On Sun, 20 Sep 2015, Zac Medico wrote: |
6 |
> > |
7 |
> >> Make the =* glob match only on boundaries between version parts, in |
8 |
> >> order to eliminate ambiguity (so that 1* does not match version |
9 |
> >> 10). Only break compatibility in cases where dependencies have been |
10 |
> >> specified ambiguously. |
11 |
> > |
12 |
> > Does this patch consider suffix parts (including the integer if |
13 |
> > present) as a single component? |
14 |
> |
15 |
> No, it considers the integer parts as a separate components, in order |
16 |
> to maintain as much backward compatibility as possible. |
17 |
> |
18 |
> So, it deviates from your test cases only because =cat/foo-1_alpha* |
19 |
> matches cat/foo-1_alpha1. Note that this case is not ambiguous like |
20 |
> the others, so we may not want to break compatibility here. |
21 |
> |
22 |
> > |
23 |
> > That is, the following should match: |
24 |
> > |
25 |
> > =cat/foo-1_alpha* cat/foo-1_alpha_beta |
26 |
> > |
27 |
> > But these should *not*: |
28 |
> > |
29 |
> > =cat/foo-1_p* cat/foo-1_pre |
30 |
> > =cat/foo-1_alpha* cat/foo-1_alpha1 |
31 |
> > |
32 |
> > (The boundary of the version components is after _alpha and _alpha1, |
33 |
> > respectively.) |
34 |
> > |
35 |
> > Ulrich |
36 |
> > |
37 |
> |
38 |
> |
39 |
|
40 |
Ok, merged approved, we can tighten it up more later easier than going |
41 |
too tight now and causing some tree havoc. |
42 |
|
43 |
|
44 |
-- |
45 |
Brian Dolbec <dolsen> |