| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
05d2fa4e71c76d7270e78ca7b1ae20a1114f07fa
|
|
|
|
|
|
| |
TAtomicIntrusivePtr
68430c6347c40eee09bc63a3441af024780b86d4
|
| |
|
|
|
|
| |
211b5254a54161c8bb4c4b077c406692e9a20cf6
|
|
|
|
|
| |
https://en.cppreference.com/w/cpp/types/numeric_limits/lowest
349097f620987be824e9db00f76af89746741c75
|
|
|
|
|
|
|
|
|
|
|
|
| |
The new format is better optimized for VectorizedRead, which is needed in Sequoia.
Previously, the output for the BeginCopy verb was a serialized subtree plus the list of opaque children. In a non-Sequoia world, that is useful and easy to work with. However, in Sequoia, a subtree might not be contained on the same cell, which makes it difficult to use the old data transfer format.
Now it's a vector of nodes with clear separation between them. This change makes it easier to work with VectorizedRead since the conversion between BeginCopy return and VectorizedRead(BeginCopy) return is very straight-forward. Additionally, this change helps with copying from Cypress to Sequoia, since now it's much easier to process each node individually (which is needed for Sequoia to work and not clog up one cell).
In the next PR, I'm planning on:
1. Making BeginCopy a read-request (removing snapshot locks). So the verb will only return meta-information about the requested nodes.
2. Adding a verb, executed before BeginCopy, where the locks would be taken and the general structure of the tree is returned.
017ec9971e8e0a611a7286ed748b6071cfc89048
|
| |
|
| |
|
|
|
|
| |
089f70efa4e50032e1c73e5c556363fa355eb4dc
|
|
|
|
| |
05b33663c7ed371efb2b290505bd8b9d8a9f1d0f
|
|
|
|
| |
3ce1f2a59c087222badce66a8c7043e77907f530
|
|
|
|
| |
e5feb0553ad2842db9974979d9808105e6c5ba34
|
|
|
|
|
| |
Moving files around so that include of either `string_builder.h` or `format.h` would result in being able to use both `Format` and `TStringBuilder`
8a2abbea2ae7027c1cd3a243ab3de55bdd5d3d27
|
|
|
|
| |
81b1c42ea62adf4f2a7e2e7ba3601ca9687e29ed
|
|
|
|
| |
7be48a0651838e770b920150e0348bf3e92416d2
|
|
|
|
| |
7788f89a156a0f6b3f6c4f4316c16e9a00096e21
|
| |
|
|
|
|
| |
6ac63b68e1f264939c9b4fc169e9381a70ceab8a
|
|
|
|
| |
467ee25e414a8d0b1f7c99d309b0eda1fb1c9932
|
|
|
|
| |
00ffbccf4bb9d61dd36ed912cde8cae38d44a858
|
|
|
|
| |
e52d229afddf7db640499d3856bc55f2b7360bb4
|
|
|
|
| |
160d1557e15ed2f25a858bb844c1543d919188a3
|
|
|
|
| |
0003a9ac951e380a13106a8f180014e0aed57f3a
|
|
|
|
| |
23b4ac85c9195fa9b0e7b0624cc3a61d010745be
|
|
|
|
| |
a27af83d265ebf4e7f4bf273eb41f69850817b05
|
|
|
|
| |
438bc5ed4e35d7a3aeffcce25e862d21289d4cad
|
|
|
|
| |
37b50713c341b80929c9c75df32b79a4e3fe00ab
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
We slightly rework inner workings of stat analysis. For almost any sane use case the behavior is identical, but should compile a little faster.
If you send string literals like `char[N]` or `const char*` make them constexpr. If you can't (e.g. `ex.what()` case) wrap it in `TStringBuf`, if you want the "by the book; 100% right" solution then wrap it in `TRuntimeFormat` instead.
`SetRequestInfo` methods now statically checks the format. Only one error was found -- hooray!
Some other internal stuff was added, a lot removed -- don't worry about it. You won't be able to see the difference
62050dfe8a9cedc1289f18e80397ff33a8e2ecdb
|
| |
|
|
|
|
| |
e856aa45df227b86e8b121852d3774bb2504193b
|
| |
|
|
|
|
| |
4ab5908e416439366466d984fc08db7254401884
|
|
|
|
| |
4792c7769dba84e38c0d981714741114add571d3
|
|
|
|
| |
810902d6e6c3104880e1ab3b34d29a8aa4bec21b
|
|
|
|
| |
e9e28de85726616f22169b34b12e1d064eea2a15
|
|
|
|
| |
d4d08d59dfff0c48a950a3faa36be4ac7e060912
|
|
|
|
|
| |
Followup for rXXXXXX - further optimize ascii-only case insensitive strings
1fca7889a074a191eadce12247bdd6dd18b75ab2
|
|
|
|
|
|
| |
library/cpp/streams/factory/open_by_signature to export
6c14e61e6b845e2aa92a25daf61de10267e9e8a0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
NYT::Format had several problems:
1. There are too many ways to enable printing of T. Not all are equally good. You could specialize TValueFormatter, you could write an overload of FormatValue, you could write an overload of ToString, you could write an overload of operator << for special stream or you could specialize the Out function.
2. If you attempt to print T which cannot be printed, you get a linker error without a proper source location which is very frustrating to work with.
3. There is no static analysis of format string performed even when it is possible.
4. If you write FormatValue overload, you still have to write ToString overload if you want to use this function (and people tend to use it quite a bit, since it is defined for util types and enums.
This pr addresses these issues to some extent. Relevant changes:
1. The only way to support NYT::Format is to define the FormatValue overload. Otherwise, you get a compile-time error.
2. Format overloads have changed: Now you have two options for general use:
```
TString Format(TStaticFormat<TArgs...> fmt, TArgs&&... args);
TString Format(TRuntimeFormat fmt, TArgs&&... args);
```
Either overload checks if TArg has a FormatValue overload. TStaticFormat performs a compile-time check of flags and the argument count. It binds to any string literal and constexpr string/string_view (and TStringBuf). TRuntimeFormat has to be mentioned explicitly. Otherwise, you will get a compile-time error for using runtime variable as a format.
3(!!!). Types which name begins with NYT:: have a specialization of ToString function which uses FormatValue. Thus, if you write class in namespace NYT and define FormatValue, you get ToString automatically. If your type is not from namespace enclosing NYT, you can just call NYT::ToString for the same effect. This limitation was caused by the fact, that we cannot review all of the external projects code which might inherit from stl classes or adopt some other questionable code practises which may completely break the dispatching mechanism of ToString due to the specialization (there were such cases). Proper documentation of this library will be added soon, so that this interaction is made known. This limitation might be lifted later
77beb68082e10aaf48be1842aad8aba63f26c1bd
|
|
|
|
| |
2c3ce3e36d35b563fe21b581380310ac84007f57
|
|
|
|
|
|
| |
На данный момент никак нельзя создать thread-safe кэш с произвольным SizeProvider, из-за того, что это не позволяет сделать шаблон `TThreadSafeCache`, при этом отредактировтаь его тоже не удастся, так как для этого нужно передать дополнительный параметр `typename TSizeProvider`, что сломает обратную совместимость, так как шаблон принимает далее переменное число аргументов (см. [TThreadSafeCache](https://a.yandex-team.ru/arcadia/library/cpp/cache/thread_safe_cache.h?rev=rXXXXXX#L15))
В связи с этим добавлен еще один хелпер, для создания LRUCache с TSizeProvider
293511a33b45f23d8afc9ff217a817481401932c
|
| |
|
| |
|
|
|
|
| |
e7cda406ffefb716562a9a7ba46607dff026f1c1
|
|
|
|
|
|
|
| |
socket by default
анонс https://at.yandex-team.ru/clubs/arcadia/30286
77f0f6dfa6c3bc8c2a8428ecf91cd76b22bdb60e
|
|
|
|
| |
78574ec3b29ba6ee7f533860ee9b27c3a91365c2
|
|
|
|
| |
f18a9a891758a7ec308b2a88b2cdf3f0942b2301
|
| |
|
|
|
|
| |
6e07ea929418b1fae4257a2af37aa0ed5799f22a
|