path: root/src/google/protobuf/compiler/javanano
AgeCommit message (Collapse)AuthorFilesLines
2014-07-15Fix access around unknownFieldData.Max Cai1-7/+2
Instead of publishing its class I chose to encapsulate the troublesome references in equals()/hashCode() in the generated code into superclass methods in ExtendableMessageNano. Changed a couple of java packages in the test suite to catch this issue easier in the future. Change-Id: I43f88411f63bb6f3ffc8d63361f2f77bebf6220a
2014-04-25Support generation of Parcelable nano messages.Jeff Davidson3-2/+20
This CL adds the "parcelable_messages" option. When enabled, all generated message classes will conform to the Android Parcelable contract. This is achieved by introducing a new parent class for generated classes which implements the required functionality. Since the store_unknown_fields option also makes use of a superclass, ExtendableMessageNano, we have two versions of the new Parcelable superclass: one extending MessageNano, and one extending ExtendableMessageNano. These classes are otherwise identical. As these classes depend on Android framework jars, they are not included in the host .jar build of the nanoproto library. Finally, add a test suite for running tests of Android-specific functionality, as this cannot be done on a desktop JVM. Change-Id: Icc2a257f03317e947f7078dbb9857c3286857497
2014-04-23Adds --ignore_service nano proto compiler flagJie Dai3-2/+13
Nano proto compiler normally throws an error if any service is defined. If --ignore-services=true is set, no error is thrown and the service is simply skipped. Change-Id: Id82583555085cc55550d03a485d3f0189885240b
2014-04-10Don't reset cachedSize to 0 in getSerializedSizeDave Hawkey1-3/+2
This avoids a race-condition when cachedSize is momentarily set to 0 for non-empty messages if multiple threads call getSerializedSize (e.g. during serialization). This is a retry of getSerializedSize() has been kept non-final so that messages generated with a previous version of the compiler will not break. Change-Id: I8d8154a10938cde579ae19c55eae55b1e70e0bda
2014-03-21Revert "Don't reset cachedSize to 0 in getSerializedSize"Wink Saville1-2/+3
This reverts commit c6e12c6702ca764486f952654ba1568f00efe813.
2014-03-21Don't reset cachedSize to 0 in getSerializedSizeDave Hawkey1-3/+2
This avoids a race-condition when cachedSize is momentarily set to 0 for non-empty messages if multiple threads call getSerializedSize (e.g. during serialization). Change-Id: I15a8ded92edbf41bf1c8d787960c5bbbc8a323c5
2014-02-17Merge "Extension overhaul."Max Cai5-112/+159
2014-01-16Fix repeated packed field merging code for non-packed data.Max Cai1-1/+4
Enum fix is already included in the previous commit. Bug: Change-Id: I9fecff3c8822918a019028eb57fa39b361a2c960
2014-01-16Add validation when parsing enum fields.Max Cai3-33/+120
Invalid values from the wire are silently ignored. Unlike full/lite, the invalid values are not stored into the unknown fields, because there's no way to get them out from Nano's unknown fields without a matching Extension. Edited README and slightly moved it towards a standalone section for Nano, independent of the Micro section. Change-Id: I2c1eb07f4d6d8f3aea242b8ddd95b9c966f3f177
2014-01-14Allow whitespace in nano codegen options.Max Cai1-22/+37
So we don't need to keep all option in a single line in the .mk files. Change-Id: I786b879b334cac4cd13b32fabcb76efe53b4ac80
2014-01-10Correctness: floating point equality using bits instead of ==.Max Cai1-40/+58
Special values for float and double make it inaccurate to test the equality with ==. The main Java library uses the standard Object.equals() implementation for all fields, which for floating point fields means Float.equals() or Double.equals(). They define equality as bitwise equality, with all NaN representations normalized to the same bit sequence (and therefore equal to each other). This test checks that the nano implementation complies with Object.equals(), so NaN == NaN and +0.0 != -0.0. Change-Id: I97bb4a3687223d8a212c70cd736436b9dd80c1d7
2014-01-10Don't serialize required fields whose 'has' flags are unset.Max Cai2-5/+7
Change-Id: Ibbe944fff83e44a8f2206e18ee9ec6f10661297a
2014-01-10Extension overhaul.Max Cai5-112/+159
- Get rid of TypeLiteral<T>. It was introduced to read the component type of a List<T> at runtime. But we use arrays everywhere else, and we can always read the component type of an array type at runtime. - Properly read/write "minor" types (e.g. sint32, sfixed32). The old implementation could only read/write data as the "typical" types (one per Java type), e.g. java.lang.Integer -> int32, java.lang.Long -> int64. So if e.g. an extension specifies sfixed32 as the type, it would be read/written in the totally incompatible int32 format. - Properly serialize repeated packed fields. The old implementation doesn't do packed serialization. As an added bonus, and to be more aligned with the rest of protobuf nano / main, repeated packable extensions can deserialize both packed and non-packed data. - Split Extension class into a hierarchy so under typical usage a large chunk of code dealing with primitive type extensions can be removed by ProGuard. Bug: Change-Id: I0d692f35cc2a8ad3a5a1cb3ce001282b2356b041
2013-12-10Avoid class initializers to help ProGuard.Max Cai10-76/+219
Class initializers prevent ProGuard from inlining any methods because it thinks the class initializer may have side effects. This is true for static methods, but instance methods can still be inlined, because to have an instance you will have touched the class and any class initializers would have run. But ProGuard only starts inlining instance methods of classes with class initializers from v4.11b6, and Android uses v4.4 now. This change tries to avoid the class initializers as much as possible, by delaying the initialization of the empty array and some fields' saved defaults until when they're needed. However, if the message hosts any extensions, they must be public static final and therefore introducing the class initializer. In that case we won't bother with lazy initialization. Change-Id: I00d8296f6eb0023112b93ee135cdb28dbd52b0b8
2013-12-09Nano: don't generate accessor methods for nested methodsAndrew Flynn3-119/+6
For nested message objects, don't generate accessor methods because they have a default value that is not a valid value (null), so there is no reason to have get/set/has/clear methods for them. Clients and protos (while serializing) can check against the invalid value to see if it's been set. Change-Id: Ic63400889581271b8cbcd9c45c84519d4921fd4b
2013-11-18Align with main: two ways of parsing repeated packable fields.Max Cai7-106/+157
It is a requirement for parsing code to handle packed and unpacked forms on the wire for repeated packable fields. This change aligns the javanano's behavior with the java's. Bonus: optimize array length calculation when parsing repeated fixed-size-element-type fields. Bonus 2: lose "xMemoizedSerializedSize" for repeated enum fields, and make the serialized size calculation match that for repeated int32 fields. Change-Id: I8a06103d9290234adb46b0971b5ed155544fe86a
2013-11-11Minimize method count for nanoproto.Tom Chao2-61/+28
- Migrates getCachedSize to the MessageNano parent class to save one method per message. - Create ExtendableMessageNano parent class for protos with extensions, this saves the getExtension and setExtension methods on the relevant messages. - getSerializedSize's default case (with no fields to serialize) also migrate to the parent class, which saves methods on empty messages. - Container classes become interfaces to save the constructor. Change-Id: I81f1a1b6d6a660096835e9df3ea20456655aab4a
2013-11-05Allow for ref-type arrays containing null elements.Aurash Mahbod2-20/+51
Strip the null elements out before serializing the array. This is helpful in the cases where the user wants to construct an array of an inexact size for serialization. For example: User constructs array of size 5 because they anticipate adding more than 1 element before serialization. Only 3 get added, so the array looks like [Obj, Obj, Obj, null, null]. This would curently crash without this CL. All repeated fields of ref-type elements can contain null elements: repeated strings, repeated bytes, and repeated messages/groups. Change-Id: I117391c868c9a436536d70d6151780e9cc7e8227 Conflicts: src/google/protobuf/compiler/javanano/
2013-10-25Implement hashCode() and equals() behind a generator option.Brian Duff13-57/+456
The option is only called 'generate_equals' because: - equals() is the main thing; hashCode() is there only to complement equals(); - it's shorter; - toString() should not be included in this option because it's more for debugging and it's more likely to stop ProGuard from working well. Also shortened the "has bit" expression; was ((bitField & mask) == mask), now ((bitField & mask) != 0). Both the Java code and the bytecode are slightly shorter. Change-Id: Ic309a08a60883bf454eb6612679aa99611620e76
2013-10-16Feature request: set() and clear() accessors return thisMax Cai3-17/+27
Also pre-inlines set() and has() in serialization code. This could theoretically help ProGuard: the message class size is usually large, and because of this only, it may refuse to inline an accessor into the serialization code, and as a result keeps the accessor intact. Chances are, after pre-inlining all accessor calls within the message class, those accessors become unused or single-use, so there are more reasons for ProGuard to inline and then remove them. Change-Id: I57decbe0b2533c1be21439de0aad15f49c7024dd
2013-10-16Make generated code more aligned with Google Java style.Max Cai6-55/+78
- Blank line after opening a message class (but not an enum interface). - Let all code blocks insert blank lines before themselves. This applies to 'package' statement, all message classes, enum classes or constant groups, extensions, bitfields, proto fields (one block per field; i.e. accessors don't have blank lines among them), and basic MessageNano methods. In this case we don't need to guess what the next block is and create blank lines for it. - Fixed some newline/indent errors. - Only one SuppressWarnings("hiding") per file. Change-Id: I865f52ad4fb6ea3b3a98b97ac9d78d19fc46c858
2013-10-15Fix repeated field merging semantics.Max Cai3-33/+48
The public doc states that repeated fields are simply concatenated and doesn't impose a different semantics for packed fields. This CL fixes this for packed fields and adds tests covering all cases. Also fixed a bit of missed null-repeated-field treatments. Change-Id: Ie35277bb1a9f0b8171dc9d07b6adf9b9d3308de2
2013-10-14Protect against null repeated fields.Brian Duff3-18/+25
There's no distinction between a repeated field being null and being empty. In both cases, nothing is sent on the wire. Clients might for whatever reason inadvertently set a repeated field to null, so protect against that and treat it just as if the field was empty. Change-Id: Ic3846f7f2189d6cfff6f8ef3ca217daecc3c8be7
2013-10-11Remove all field initializers and let ctor call clear().Max Cai4-14/+18
The field initializers have basically caused the compiled <init> method to inline the whole clear() method, which means if ProGuard is not used or failed to inline or remove clear(), there are two big chunks of code that do the same thing. So why not just call clear() from the ctor. Change-Id: Ief71e2b03db2e059b3bfa98309649368089ffab0
2013-10-09Merge "Add reftypes field generator option."Ulas Kirazci5-10/+51
2013-10-08Fix some indenting issues with set__() functionAndrew Flynn1-10/+6
Previously it looked like this: public final class OuterClass { [...] public static final class InnerClass extends { [...] public void setId(java.lang.String value) { if (value == null) { throw new java.lang.NullPointerException(); } id_ = value; bitfield0_ |= 0x00000001; [...] } [...] } Now it looks like this: public final class OuterClass { [...] public static final class InnerClass extends { [...] public void setId(java.lang.String value) { if (value == null) throw new java.lang.NullPointerException(); id_ = value; bitfield0_ |= 0x00000001; [...] } [...] } Change-Id: I2a9289b528f785c846210d558206d677aa13e9be
2013-10-07Add reftypes field generator option.Brian Duff5-10/+51
This option generates fields as reference types, and serializes based on nullness. Change-Id: Ic32e0eebff59d14016cc9a19e15a9bb08ae0bba5 Signed-off-by: Brian Duff <>
2013-09-25Implement enum_style=java option.Max Cai3-19/+65
This javanano_out command line option creates a container interface at the normal place where the enum constants would reside, per enum definition. The java_multiple_files flag would now affect the file- scope enums with the shells. If the flag is true then file-scope container interfaces are created in their own files. Change-Id: Id52258fcff8d3dee9db8f3d8022147a811bf3565
2013-09-23Accessor style for optional fields.Max Cai11-56/+409
This CL implements the 'optional_field_style=accessors' option. All optional fields will now be 1 Java field and 1 bit in a shared bitfield behind get/set/has/clear accessor methods. The setter performs null check for reference types (Strings and byte[]s). Also decentralized the clear code generation. Change-Id: I60ac78329e352e76c2f8139fba1f292383080ad3
2013-09-20Add some bitfield helper methods from 2.4Max Cai2-0/+102
Change-Id: Ib9bb549602f71a451d2107fb04de17877553860e
2013-09-18Add two codegen parameters to nano.Max Cai2-2/+30
enum_style = c | java: 'c' to put the enum member int constants at the parent scope; 'java' to create uninstantiatable shell classes at the parent scope and put the int constants inside. optional_field_style = default | accessors: 'default' to create one public mutable field per optional proto field; 'accessors' to encapsulate the generated fields behind get, set, has and clear accessors. This CL only contains parsing code for these two parameters. Change-Id: Iec0c3b0f30af8eb7db328e790664306bc90be089
2013-08-07Merge "Fix outer classname for javamicro/javanano."Wink Saville6-207/+110
2013-08-05Fix outer classname for javamicro/javanano.Max Cai6-207/+110
- File class name is defined as the java_outer_classname option value or the file name ToCamelCase; never the single message's ClassName. - File-scope enums are translated to constants in the file class, regardless of java_multiple_files. - If java_multiple_files=true, and file's class name equals a message's class name, no error. This is done by detecting that the outer class is not needed and skipping the outer class codegen and clash checks. Note: there is a disparity between java[lite] and the previous java{micr|nan}o: when generating code for a single-message proto, the outer class is omitted by java{micr|nan}o if the file does not have java_outer_classname. This change makes java{micr|nan}o align with java[lite] codegen and create the outer class, but will print some info to warn of potential change of code. - Also fixed the "is_own_file" detection and made all parseX() methods static. Previously, all messages in a java_multiple_files=true file are (incorrectly) considered to be in their own files, including nested messages, causing them to become inner classes (instance- bound) and forcing the parseX() methods to lose the static modifier. - This change supersedes c/60164 and c/60086, which causes javanano to put enum values into enum shell classes if java_multiple_files=true. We now always use the parent class to host the enum values. A future change will add a command line option to provide more flexibility. - Elaborated in java/README.txt. Change-Id: I684932f90e0a028ef37c662b221def5ffa202439
2013-07-31Merge "Fixed packed repeated serialization."Ulas Kirazci2-30/+29
2013-07-31Fixed packed repeated serialization.Ulas Kirazci2-30/+29
Remove buggy memoization. Memoization also is too fragile for the api because the repeated field is public. Change-Id: I538b8426d274b22df2eeea5935023abbe7df49fe
2013-07-29Add an option to inspect "has" state upon parse.Ulas Kirazci6-31/+83
If has is set, also always serialize. Change-Id: I2c8450f7ab9e837d722123dd1042991c0258ede3
2013-07-25Per-file java_multiple_files flag.Max Cai6-24/+54
Imported source files may have different values for the 'java_multiple_files' option to the main source file's. Whether the fully qualified Java name of an entity should include the outer class name depends on the flag value in the file defining the referenced entity, not the main file. This CL loads the flag values from the main and all transitively imported files into the params, and generates the fully qualified Java names accordingly. If the generator option 'java_multiple_files' is set, its value overrides any in-file values in all source/imported files. This is because this generator option is typically used on either none or all source files. Change-Id: Id6a4a42426d68961dc669487d38f35530deb7d8e
2013-07-23Allow NaN/+inf/-inf defaults in micro/nano.Chris Smith2-4/+64
Adds support for default values of NaN, infinity and negative infinity for floats and doubles in both the nano and micro java compiler. Change-Id: Ibc43e5ebb073e51d9a8181f3aa23b72e10015dca
2013-07-19Update nano to serialize java keywords properly.Tom Chao8-11/+63
Change-Id: I7407d0fab609c336ecd73499e725aed0dd50f555
2013-07-08Make it possible to use MessageNano.mergeFrom without casting.Brian Duff1-1/+1
You can now do: MyMessage foo = MessageNano.mergeFrom(new MyMessage(), bytes); without having to cast the message returned from mergeFrom. Change-Id: Ibb2ad327f75855d45352ad304c7f054f20dd29c9
2013-06-24Nano support for extensions and unknown fields.Brian Duff6-21/+264
You can use the processor option store_unknown_fields to switch this support on: aprotoc --javanano_out=store_unknown_fields=true:/tmp/out A separate option for extensions isn't required. Support for unknown fields must be turned on to allow storing and retrieving extensions, because they are just stored as unknown fields. If unknown fields are switched on, extension related code will be generated when a proto message includes an extension range, or an extension is encountered. By default, store_unknown_fields is false. No additional code is generated, and the generator will error out if protos contain extension ranges or extensions. Change-Id: I1e034c9e8f3305612953f72438189a7da6ed2167
2013-06-07Fix enum field references with java_multiple_files.Brian Duff1-4/+14
When the java_multiple_files option is on, enums are placed in java class files based on the name of the original enum type. This fixes field references to such enum values to point to the correct class name when setting the default. Change-Id: I51a2e251c0d0ab1e45a182ba849d314232a74bac
2013-06-05Fix javanano compiler generating uncompileable java code. This happenedBrian Duff1-0/+15
for enums when java_multiple_files=true. Change-Id: Ia6fe1a51c6a46eb9a2f29527829794076b165cb7 Signed-off-by: Brian Duff <>
2013-04-08Don't use Arrays.copyOf in generated code.Ficus Kirkpatrick3-4/+20
It didn't appear until API 9 and is thus incompatible with Froyo. Instead, allocate a new array and System.arraycopy inline. Change-Id: I2e1cd07a4a762ef8edd5ec06ceaa1d38b302823d
2013-04-03Prevent conflicts between local and proto variable names.Ulas Kirazci7-118/+63
Prefix access to proto variable names with "this.". Also remove unused GenerateMergingCode. Change-Id: I5f07d3252fc385c4174d9165b64785b40f676e17
2013-04-01Nano protobufs.Ulas Kirazci19-0/+3351
Like micro protobufs except: - No setter/getter/hazzer functions. - Has state is not available. Outputs all fields != their default. - CodedInputStream can only take byte[] (not InputStream). - Repeated fields are in arrays, not ArrayList or Vector. - Unset messages/groups are null, not "defaultInstance()". - Required fields are always serialized. To use: - Link libprotobuf-java-2.3.0-nano runtime. - Use LOCAL_PROTOC_OPTIMIZE_TYPE := nano Change-Id: I7429015b3c5f7f38b7be01eb2d4927f7a9999c80