You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@thrift.apache.org by "Spiridon Eliopoulos (JIRA)" <ji...@apache.org> on 2010/05/22 18:44:16 UTC
[jira] Created: (THRIFT-787) Enums are not read correctly
Enums are not read correctly
----------------------------
Key: THRIFT-787
URL: https://issues.apache.org/jira/browse/THRIFT-787
Project: Thrift
Issue Type: Bug
Components: Compiler (Haskell)
Affects Versions: 0.1
Environment: All platforms
Reporter: Spiridon Eliopoulos
Assignee: Spiridon Eliopoulos
Priority: Minor
The Haskell (and OCaml) generators emit incorrect code for enum deserilalization. The value counter is not reset to -1 before the deserialization function is generated.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (THRIFT-787) Enums are not read correctly
Posted by "Christian Lavoie (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/THRIFT-787?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12915455#action_12915455 ]
Christian Lavoie commented on THRIFT-787:
-----------------------------------------
I generated code with all generators, using a simplistic enum:
{code}
enum MyEnum {
ZERO,
ONE,
THREE = 3,
FOUR,
FORTYTWO = 42,
FORTYTHREE,
TWELVE = 12,
THIRTEEN
}
{code}
{noformat}
9050e1c444bb9ec823857e68f1ea3c6e tmp-without-787/gen-as3/MyEnum.as
e364dd1d37f48f5ee49898273d583ac3 tmp-without-787/gen-cocoa/test.h
13353d093c7fbfb6d34e18c5e209834e tmp-without-787/gen-cocoa/test.m
0e35d18e6b3d161050416ef65f589a7b tmp-without-787/gen-cpp/test_constants.cpp
0305410c79ead19f7a21da83005bbc4d tmp-without-787/gen-cpp/test_constants.h
d614d268bf718712f9857febe59a87a5 tmp-without-787/gen-cpp/test_types.cpp
f4202dd6e56eb234e9ff319fd5248b58 tmp-without-787/gen-cpp/test_types.h
44537aa869bdf2c22d2a40427eb7bfbb tmp-without-787/gen-csharp/MyEnum.cs
5cc5162f68c30a8ff998f0ba28afd1ec tmp-without-787/gen-erl/test_constants.hrl
01ea91d8c4d3ba7fe0c4568aecf66636 tmp-without-787/gen-erl/test_types.erl
80fcb1e546073b806fc1676176330477 tmp-without-787/gen-erl/test_types.hrl
515131db34af21096383eadc718905ab tmp-without-787/gen-hs/Test_Consts.hs
c65c348912562e58a25d4448b5881f06 tmp-without-787/gen-hs/Test_Types.hs
7de88e95d2037748dd0f3917b5645265 tmp-without-787/gen-html/index.html
c6e33f8d8d4b88f37d666ed8392d24cb tmp-without-787/gen-html/style.css
6d86e7f688a70ab75289fbfe14f67f39 tmp-without-787/gen-html/test.html
53183bfed8fb46399da51ad849758e4e tmp-without-787/gen-java/MyEnum.java
1f32e388a639a566394a1f6611cebea6 tmp-without-787/gen-javame/MyEnum.java
ae31c442df8f2982efbd3bbcbf6c6ddd tmp-without-787/gen-js/test_types.js
ad2f732e3f82c4d02baf90b5800ed4b0 tmp-without-787/gen-ocaml/test_consts.ml
27a042e3e8a036dbcb374a8103ea7d35 tmp-without-787/gen-ocaml/test_types.ml
0e2fe71b03b241ff4c2b5c1a4b36e208 tmp-without-787/gen-ocaml/test_types.mli
de37caf4bb38ca4399b31850dddcd0ec tmp-without-787/gen-perl/Constants.pm
07be59e6b0ce725187869ebc1d2c9a14 tmp-without-787/gen-perl/Types.pm
fd9fd449f522a983f545d3dbb038ad4b tmp-without-787/gen-php/test/test_types.php
d41d8cd98f00b204e9800998ecf8427e tmp-without-787/gen-py/__init__.py
fcb6376ae0410613e38952e6b0395814 tmp-without-787/gen-py/test/__init__.py
14e2d529821f06f285353a86634e673b tmp-without-787/gen-py/test/constants.py
94c4a93659d81bc9b48ad2519ea9a37f tmp-without-787/gen-py/test/ttypes.py
8096a712cd33880d33dca92b11671472 tmp-without-787/gen-rb/test_constants.rb
28a82f55b1a1d74e5e7882292f1eab8b tmp-without-787/gen-rb/test_types.rb
76c2d5049d0c0b0df90cec6760815ecb tmp-without-787/gen-st/test.st
718d8596a14d123d83afb0d5d6d6fd96 tmp-without-787/gen-xsd/test_xsd.php
{noformat}
Then again, after apply my v2 patch:
{noformat}
9050e1c444bb9ec823857e68f1ea3c6e tmp-with-787/gen-as3/MyEnum.as
e364dd1d37f48f5ee49898273d583ac3 tmp-with-787/gen-cocoa/test.h
13353d093c7fbfb6d34e18c5e209834e tmp-with-787/gen-cocoa/test.m
0e35d18e6b3d161050416ef65f589a7b tmp-with-787/gen-cpp/test_constants.cpp
0305410c79ead19f7a21da83005bbc4d tmp-with-787/gen-cpp/test_constants.h
d614d268bf718712f9857febe59a87a5 tmp-with-787/gen-cpp/test_types.cpp
f4202dd6e56eb234e9ff319fd5248b58 tmp-with-787/gen-cpp/test_types.h
44537aa869bdf2c22d2a40427eb7bfbb tmp-with-787/gen-csharp/MyEnum.cs
5cc5162f68c30a8ff998f0ba28afd1ec tmp-with-787/gen-erl/test_constants.hrl
01ea91d8c4d3ba7fe0c4568aecf66636 tmp-with-787/gen-erl/test_types.erl
80fcb1e546073b806fc1676176330477 tmp-with-787/gen-erl/test_types.hrl
515131db34af21096383eadc718905ab tmp-with-787/gen-hs/Test_Consts.hs
c65c348912562e58a25d4448b5881f06 tmp-with-787/gen-hs/Test_Types.hs
7de88e95d2037748dd0f3917b5645265 tmp-with-787/gen-html/index.html
c6e33f8d8d4b88f37d666ed8392d24cb tmp-with-787/gen-html/style.css
6d86e7f688a70ab75289fbfe14f67f39 tmp-with-787/gen-html/test.html
53183bfed8fb46399da51ad849758e4e tmp-with-787/gen-java/MyEnum.java
1f32e388a639a566394a1f6611cebea6 tmp-with-787/gen-javame/MyEnum.java
ae31c442df8f2982efbd3bbcbf6c6ddd tmp-with-787/gen-js/test_types.js
ad2f732e3f82c4d02baf90b5800ed4b0 tmp-with-787/gen-ocaml/test_consts.ml
27a042e3e8a036dbcb374a8103ea7d35 tmp-with-787/gen-ocaml/test_types.ml
0e2fe71b03b241ff4c2b5c1a4b36e208 tmp-with-787/gen-ocaml/test_types.mli
de37caf4bb38ca4399b31850dddcd0ec tmp-with-787/gen-perl/Constants.pm
07be59e6b0ce725187869ebc1d2c9a14 tmp-with-787/gen-perl/Types.pm
fd9fd449f522a983f545d3dbb038ad4b tmp-with-787/gen-php/test/test_types.php
d41d8cd98f00b204e9800998ecf8427e tmp-with-787/gen-py/__init__.py
fcb6376ae0410613e38952e6b0395814 tmp-with-787/gen-py/test/__init__.py
14e2d529821f06f285353a86634e673b tmp-with-787/gen-py/test/constants.py
94c4a93659d81bc9b48ad2519ea9a37f tmp-with-787/gen-py/test/ttypes.py
8096a712cd33880d33dca92b11671472 tmp-with-787/gen-rb/test_constants.rb
28a82f55b1a1d74e5e7882292f1eab8b tmp-with-787/gen-rb/test_types.rb
e74e255a131e643e8a887efcebe31156 tmp-with-787/gen-st/test.st
718d8596a14d123d83afb0d5d6d6fd96 tmp-with-787/gen-xsd/test_xsd.php
{noformat}
They're the same files, so the patch is clearly a no-op from that perspective. Also, I read through the generated code, it is "obviously correct", meaning that {{ZERO}} matches with {{0}}, {{THIRTEEN}} with {{13}}, ... in all generated code. Diff'ing leads to the same conclusion:
{noformat}
$ diff --recursive tmp-with*
diff --recursive tmp-with-787/gen-st/test.st tmp-without-787/gen-st/test.st
15c15
< !Test class methodsFor: 'as yet uncategorized' stamp: 'thrift 09/27/2010 15:57'!
---
> !Test class methodsFor: 'as yet uncategorized' stamp: 'thrift 09/27/2010 15:54'!
19c19
< !Test class methodsFor: 'as yet uncategorized' stamp: 'thrift 09/27/2010 15:57'!
---
> !Test class methodsFor: 'as yet uncategorized' stamp: 'thrift 09/27/2010 15:54'!
23c23
< !Test class methodsFor: 'as yet uncategorized' stamp: 'thrift 09/27/2010 15:57'!
---
> !Test class methodsFor: 'as yet uncategorized' stamp: 'thrift 09/27/2010 15:54'!
27c27
< !Test class methodsFor: 'as yet uncategorized' stamp: 'thrift 09/27/2010 15:57'!
---
> !Test class methodsFor: 'as yet uncategorized' stamp: 'thrift 09/27/2010 15:54'!
{noformat}
The only difference is some timestamp in the st generator.
> Enums are not read correctly
> ----------------------------
>
> Key: THRIFT-787
> URL: https://issues.apache.org/jira/browse/THRIFT-787
> Project: Thrift
> Issue Type: Bug
> Components: Haskell - Compiler
> Affects Versions: 0.1
> Environment: All platforms
> Reporter: Spiridon Eliopoulos
> Assignee: Spiridon Eliopoulos
> Priority: Minor
> Fix For: 0.5
>
> Attachments: 0001--THRIFT-787-Fix-bug-in-enum-deserialization.patch, v2-thrift-read-enums-cleanup.patch
>
>
> The Haskell (and OCaml) generators emit incorrect code for enum deserilalization. The value counter is not reset to -1 before the deserialization function is generated.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (THRIFT-787) Enums are not read correctly
Posted by "Bryan Duxbury (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/THRIFT-787?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Bryan Duxbury updated THRIFT-787:
---------------------------------
Fix Version/s: 0.5
> Enums are not read correctly
> ----------------------------
>
> Key: THRIFT-787
> URL: https://issues.apache.org/jira/browse/THRIFT-787
> Project: Thrift
> Issue Type: Bug
> Components: Haskell - Compiler
> Affects Versions: 0.1
> Environment: All platforms
> Reporter: Spiridon Eliopoulos
> Assignee: Spiridon Eliopoulos
> Priority: Minor
> Fix For: 0.5
>
> Attachments: 0001--THRIFT-787-Fix-bug-in-enum-deserialization.patch, v2-thrift-read-enums-cleanup.patch
>
>
> The Haskell (and OCaml) generators emit incorrect code for enum deserilalization. The value counter is not reset to -1 before the deserialization function is generated.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Issue Comment Edited: (THRIFT-787) Enums are not read
correctly
Posted by "Christian Lavoie (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/THRIFT-787?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12914898#action_12914898 ]
Christian Lavoie edited comment on THRIFT-787 at 9/25/10 9:20 PM:
------------------------------------------------------------------
Actually, this is now handled in parse/t_enum.*, the resolve_values call. It's implemented redundantly in (as far as I can tell) all generators the same way:
{noformat}
286 for (c_iter = constants.begin(); c_iter != constants.end(); ++c_iter) {
287 if ((*c_iter)->has_value()) {
288 value = (*c_iter)->get_value();
289 } else {
290 ++value; // THIS WILL NEVER EXECUTE
291 }
292 string name = capitalize((*c_iter)->get_name());
293
294 f_types_ <<
295 indent() << name << " -> " << value << endl;
296 }
{noformat}
(or the appropriate variant thereof)
So this patch is now redundant (thanks to resolve_values, the else branch of that if statement is never taken in the generators). Cleanup patch incoming as soon as I have a setup where I can run tests.
was (Author: clavoie):
Actually, this is now handled in parse/t_enum.*, the resolve_values call. It's implemented redundantly in (as far as I can tell) all generators the same way:
286 for (c_iter = constants.begin(); c_iter != constants.end(); ++c_iter) {
287 if ((*c_iter)->has_value()) {
288 value = (*c_iter)->get_value();
289 } else {
290 ++value; // THIS WILL NEVER EXECUTE
291 }
292 string name = capitalize((*c_iter)->get_name());
293
294 f_types_ <<
295 indent() << name << " -> " << value << endl;
296 }
(or the appropriate variant thereof)
So this patch is now redundant (thanks to resolve_values, the else branch of that if statement is never taken in the generators). Cleanup patch incoming as soon as I have a setup where I can run tests.
> Enums are not read correctly
> ----------------------------
>
> Key: THRIFT-787
> URL: https://issues.apache.org/jira/browse/THRIFT-787
> Project: Thrift
> Issue Type: Bug
> Components: Haskell - Compiler
> Affects Versions: 0.1
> Environment: All platforms
> Reporter: Spiridon Eliopoulos
> Assignee: Spiridon Eliopoulos
> Priority: Minor
> Attachments: 0001--THRIFT-787-Fix-bug-in-enum-deserialization.patch
>
>
> The Haskell (and OCaml) generators emit incorrect code for enum deserilalization. The value counter is not reset to -1 before the deserialization function is generated.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (THRIFT-787) Enums are not read correctly
Posted by "Christian Lavoie (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/THRIFT-787?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12914898#action_12914898 ]
Christian Lavoie commented on THRIFT-787:
-----------------------------------------
Actually, this is now handled in parse/t_enum.*, the resolve_values call. It's implemented redundantly in (as far as I can tell) all generators the same way:
286 for (c_iter = constants.begin(); c_iter != constants.end(); ++c_iter) {
287 if ((*c_iter)->has_value()) {
288 value = (*c_iter)->get_value();
289 } else {
290 ++value; // THIS WILL NEVER EXECUTE
291 }
292 string name = capitalize((*c_iter)->get_name());
293
294 f_types_ <<
295 indent() << name << " -> " << value << endl;
296 }
(or the appropriate variant thereof)
So this patch is now redundant (thanks to resolve_values, the else branch of that if statement is never taken in the generators). Cleanup patch incoming as soon as I have a setup where I can run tests.
> Enums are not read correctly
> ----------------------------
>
> Key: THRIFT-787
> URL: https://issues.apache.org/jira/browse/THRIFT-787
> Project: Thrift
> Issue Type: Bug
> Components: Haskell - Compiler
> Affects Versions: 0.1
> Environment: All platforms
> Reporter: Spiridon Eliopoulos
> Assignee: Spiridon Eliopoulos
> Priority: Minor
> Attachments: 0001--THRIFT-787-Fix-bug-in-enum-deserialization.patch
>
>
> The Haskell (and OCaml) generators emit incorrect code for enum deserilalization. The value counter is not reset to -1 before the deserialization function is generated.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Closed: (THRIFT-787) Enums are not read correctly
Posted by "Bryan Duxbury (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/THRIFT-787?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Bryan Duxbury closed THRIFT-787.
--------------------------------
Assignee: Christian Lavoie (was: Spiridon Eliopoulos)
Resolution: Fixed
I just committed this. There were a few other little blobs of calls to the has_value() method in as3, python, and cpp compilers that seemed superfluous, so I whacked those too. Thanks for the patch, Christian!
> Enums are not read correctly
> ----------------------------
>
> Key: THRIFT-787
> URL: https://issues.apache.org/jira/browse/THRIFT-787
> Project: Thrift
> Issue Type: Bug
> Components: Haskell - Compiler
> Affects Versions: 0.1
> Environment: All platforms
> Reporter: Spiridon Eliopoulos
> Assignee: Christian Lavoie
> Priority: Minor
> Fix For: 0.5
>
> Attachments: 0001--THRIFT-787-Fix-bug-in-enum-deserialization.patch, v2-thrift-read-enums-cleanup.patch
>
>
> The Haskell (and OCaml) generators emit incorrect code for enum deserilalization. The value counter is not reset to -1 before the deserialization function is generated.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (THRIFT-787) Enums are not read correctly
Posted by "Spiridon Eliopoulos (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/THRIFT-787?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Spiridon Eliopoulos updated THRIFT-787:
---------------------------------------
Attachment: 0001--THRIFT-787-Fix-bug-in-enum-deserialization.patch
This patch fixes this issue.
> Enums are not read correctly
> ----------------------------
>
> Key: THRIFT-787
> URL: https://issues.apache.org/jira/browse/THRIFT-787
> Project: Thrift
> Issue Type: Bug
> Components: Compiler (Haskell)
> Affects Versions: 0.1
> Environment: All platforms
> Reporter: Spiridon Eliopoulos
> Assignee: Spiridon Eliopoulos
> Priority: Minor
> Attachments: 0001--THRIFT-787-Fix-bug-in-enum-deserialization.patch
>
>
> The Haskell (and OCaml) generators emit incorrect code for enum deserilalization. The value counter is not reset to -1 before the deserialization function is generated.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (THRIFT-787) Enums are not read correctly
Posted by "Christian Lavoie (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/THRIFT-787?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Christian Lavoie updated THRIFT-787:
------------------------------------
Attachment: v2-thrift-read-enums-cleanup.patch
Haven't tested this yet, but reading the code produced by the compiler, it all looks 'obviously correct' to me (keep in mind I don't speak as3, cocoa, csharp, erl, php, st or xsd fluently).
I should be able test about half the generators first hand this week, if nobody beats me to it.
> Enums are not read correctly
> ----------------------------
>
> Key: THRIFT-787
> URL: https://issues.apache.org/jira/browse/THRIFT-787
> Project: Thrift
> Issue Type: Bug
> Components: Haskell - Compiler
> Affects Versions: 0.1
> Environment: All platforms
> Reporter: Spiridon Eliopoulos
> Assignee: Spiridon Eliopoulos
> Priority: Minor
> Attachments: 0001--THRIFT-787-Fix-bug-in-enum-deserialization.patch, v2-thrift-read-enums-cleanup.patch
>
>
> The Haskell (and OCaml) generators emit incorrect code for enum deserilalization. The value counter is not reset to -1 before the deserialization function is generated.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.