contracts: Convert to framev2 macros

[athei]

Convert to the new frame macros.

Metadata diff:

diff --git a/old.json b/new.json
index 2201706..0467169 100644
--- a/old.json
+++ b/new.json
@@ -744,7 +744,7 @@
         "name": "DeletionQueue",
         "modifier": "Default",
         "ty": {
-          "Plain": "Vec<storage::DeletedContract>"
+          "Plain": "Vec<DeletedContract>"
         },
         "default": [
           0
@@ -786,7 +786,7 @@
         },
         {
           "name": "gas_limit",
-          "ty": "Compact<Gas>"
+          "ty": "Compact<Weight>"
         },
         {
           "name": "data",
@@ -812,7 +812,7 @@
         },
         {
           "name": "gas_limit",
-          "ty": "Compact<Gas>"
+          "ty": "Compact<Weight>"
         },
         {
           "name": "code",
@@ -860,7 +860,7 @@
         },
         {
           "name": "gas_limit",
-          "ty": "Compact<Gas>"
+          "ty": "Compact<Weight>"
         },
         {
           "name": "code_hash",
@@ -1206,8 +1206,7 @@
         0
       ],
       "documentation": [
-        " The maximum nesting level of a call/instantiate stack. A reasonable default",
-        " value is 100."
+        " The maximum nesting level of a call/instantiate stack."
       ]
     },
     {
@@ -1220,7 +1219,7 @@
         0
       ],
       "documentation": [
-        " The maximum size of a storage value in bytes. A reasonable default is 16 KiB."
+        " The maximum size of a storage value and event payload in bytes."
       ]
     },
     {
@@ -1252,6 +1251,21 @@
       "documentation": [
         " The maximum amount of weight that can be consumed per block for lazy trie removal."
       ]
+    },
+    {
+      "name": "MaxCodeSize",
+      "ty": "u32",
+      "value": [
+        0,
+        0,
+        2,
+        0
+      ],
+      "documentation": [
+        " The maximum length of a contract code in bytes. This limit applies to the instrumented",
+        " version of the code. Therefore `instantiate_with_code` can fail even when supplying",
+        " a wasm binary below this maximum size."
+      ]
     }
   ],
   "errors": [
@@ -1423,7 +1437,7 @@
       "documentation": [
         " Removal of a contract failed because the deletion queue is full.",
         "",
-        " This can happen when either calling [`Module::claim_surcharge`] or `seal_terminate`.",
+        " This can happen when either calling [`Pallet::claim_surcharge`] or `seal_terminate`.",
         " The queue is filled by deleting contracts and emptied by a fixed amount each block.",
         " Trying again during another block is the only way to resolve this issue."
       ]
@@ -1433,7 +1447,7 @@
       "documentation": [
         " A contract could not be evicted because it has enough balance to pay rent.",
         "",
-        " This can be returned from [`Module::claim_surcharge`] because the target",
+        " This can be returned from [`Pallet::claim_surcharge`] because the target",
         " contract has enough balance to pay for its rent."
       ]
     },

Reduce API surface

Remove some types from the public API that does not need to be exposed. This gives us more leeway to make changes without bumping the major version once v3 is out.

This also seals the Ext trait which is not meant to be implemented by downstream crates. Implementations of it are merely passed to the chain extension for consumption.

[gui1117]

usually I like to warn about the breaking change: pallet prefix (i.e. the prefix used by storages) is now the name of the pallet given to construct_runtime instead of "Contracts":
Thus we can a runtime-migration flag IMHO, and put this comment in the top comment

#8151 (comment)

[ascjones]

LGTM 🚀

[athei]

usually I like to warn about the breaking change: pallet prefix (i.e. the prefix used by storages) is now the name of the pallet given to construct_runtime instead of "Contracts":
Thus we can a runtime-migration flag IMHO, and put this comment in the top comment

#8151 (comment)

The thing is that almost all changes leading up from v2 to v3 are breaking and would need a storage migration. The thing is that we just do not care what happens to compatibility pre v3. Doing that for this single PR would be misleading?

[gui1117]

The thing is that almost all changes leading up from v2 to v3 are breaking and would need a storage migration. The thing is that we just do not care what happens to compatibility pre v3. Doing that for this single PR would be misleading?

yes ok that makes sense, not to do for this single PR

[athei]

bot merge

[ghost]

Waiting for commit status.